---
title: Lie Algebra of Biderivations
url: https://www.emergentmind.com/topics/lie-algebra-of-biderivations
type: topic
---

# Lie Algebra of Biderivations

Searching arXiv for recent and foundational papers on Lie-algebra biderivations to ground the article in published work.
A biderivation on a Lie algebra is a bilinear map that is a derivation in each argument separately. In the adjoint-valued case, if \(L\) is a Lie algebra and \(D:L\times L\to L\), this means
\[
D([x,y],z)=[D(x,z),y]+[x,D(y,z)],\qquad
D(x,[y,z])=[D(x,y),z]+[y,D(x,z)].
\]
The subject has developed along two main lines: structural classification of biderivations on specific Lie algebras and analysis of the algebraic structures they induce, including commuting maps, commutative post-Lie or ABD-structures, and, more recently, Lie brackets on suitable spaces of one-sided biderivations. For finite-dimensional complex simple Lie algebras, the central result is a rigidity theorem: every biderivation is inner, i.e. a scalar multiple of the Lie bracket, with no skew-symmetry assumption required [1610.03765]. Subsequent work extended this picture to semisimple and complete Lie algebras [2307.15750], clarified the vanishing of symmetric biderivations in semisimple and perfect settings [2407.05581], [2503.13059], and identified families of non-inner, often symmetric, biderivations in non-simple, infinite-dimensional, or centrally extended contexts such as \(\mathfrak{gl}_n\), Schrödinger–Virasoro algebras, and \(W(a,b)\) [1610.03765], [1611.04082], [1712.09202].

## 1. Definition and formal framework

A biderivation of a Lie algebra \(L\) over a field is a bilinear map whose partial maps are derivations. In module-valued form, if \(M\) is an \(L\)-module and \(\phi:L\times L\to M\), then \(\phi\) is a biderivation when
\[
\phi([x,y],z)=x\cdot \phi(y,z)-y\cdot \phi(x,z),\qquad
\phi(x,[y,z])=y\cdot \phi(x,z)-z\cdot \phi(x,y).
\]
For the adjoint module \(M=L\), these identities become
\[
\phi([x,y],z)=[x,\phi(y,z)]-[y,\phi(x,z)],\qquad
\phi(x,[y,z])=[y,\phi(x,z)]-[z,\phi(x,y)],
\]
which is the formulation used throughout the Lie-algebra literature on the topic [2503.13059], [2407.05581].

Two symmetry types are standard. A biderivation is **symmetric** if \(\phi(x,y)=\phi(y,x)\), and **skew-symmetric** if \(\phi(x,y)=-\phi(y,x)\). Much of the earlier literature concentrated on the skew-symmetric case, especially because of its relationship with centroidal maps and commuting mappings on perfect centerless Lie algebras [1801.01109]. Later work removed the skew-symmetry restriction in several important settings, most notably for finite-dimensional complex simple Lie algebras [1610.03765].

For complete Lie algebras, meaning \(Z(L)=0\) and \(\operatorname{Der}(L)=\operatorname{ad}(L)\), any biderivation can be written in “inner-in-each-slot” form
\[
B(x,y)=[\varphi(x),y]=[x,\psi(y)]
\]
for suitable endomorphisms \(\varphi,\psi\) [2307.15750]. In the simple case this representation collapses further because the centroid is \(\mathbb{C}\,\mathrm{id}\), forcing scalar multiples of the bracket [1610.03765].

A distinct but related development considers **right** and **left** biderivations separately. A right biderivation requires only the derivation identity in the first variable, and a left biderivation only in the second. Two-sided biderivations are precisely the intersection of these classes [2507.12089]. This one-sided perspective is the setting in which an explicit Lie bracket on spaces of biderivations has recently been introduced [2507.12089].

## 2. Rigidity on simple, semisimple, and complete Lie algebras

The foundational rigidity theorem is due to Tang: if \(L\) is a finite-dimensional complex simple Lie algebra, then every biderivation \(D\) of \(L\) is inner in the sense
\[
D(x,y)=\lambda[x,y]\qquad (\lambda\in\mathbb{C}),
\]
and no skew-symmetry assumption is needed [1610.03765]. The proof proceeds by fixing one variable, observing that each partial map is a derivation, and using \(\operatorname{Der}(L)=\operatorname{ad}(L)\) to obtain linear maps \(\phi,\psi\) with
\[
D(x,y)=[\phi(x),y]=[x,\psi(y)].
\]
A Cartan subalgebra and root-space decomposition then force \(\phi\) and \(\psi\) to be scalar on each root space; connectivity of the root system shows the scalar is constant across all roots; and the Cartan part follows by comparing brackets with root vectors [1610.03765]. The result applies uniformly to all finite-dimensional complex simple Lie algebras, including classical and exceptional types [1610.03765].

This theorem admits a direct semisimple extension. If
\[
L=\bigoplus_{i=1}^t L_i
\]
is a finite-dimensional complex semisimple Lie algebra with simple ideals \(L_i\), then every biderivation is of the form
\[
B(x,y)=\sum_{i=1}^t \lambda_i [x_i,y_i],
\]
where \(x=\sum_i x_i\), \(y=\sum_i y_i\), and each \(\lambda_i\in\mathbb{C}\) acts on a simple summand [2307.15750]. In this sense,
\[
\operatorname{BiDer}(L)\cong \operatorname{Cent}(L)\cong \mathbb{C}^t
\]
for semisimple \(L\) [2307.15750].

The same paper places these results in the broader context of complete Lie algebras. For complete \(L\), every biderivation is governed by endomorphisms \(\varphi,\psi\) satisfying \([\varphi(x),y]=[x,\psi(y)]\), and a matrix formalism expresses this through relations
\[
PA_i=A_iQ
\]
for the structure matrices \(A_i\) of \(L\) [2307.15750]. In semisimple cases this collapses to scalar block-diagonal solutions and reproduces the classification above [2307.15750].

A parallel route to rigidity comes from the skew-symmetric theory of Brešar and Zhao. If \(L\) is perfect and centerless, then every skew-symmetric biderivation is of the form
\[
\delta(x,y)=\gamma([x,y])
\]
for some \(\gamma\in \operatorname{Cent}(L)\) [1801.01109]. On simple Lie algebras the centroid is scalar, so this again yields \(\delta(x,y)=\lambda[x,y]\) [1801.01109]. Tang’s theorem strengthens this in the simple case by showing that the same conclusion holds even without skew-symmetry [1610.03765].

## 3. Symmetric biderivations and vanishing theorems

Symmetric biderivations behave very differently from skew-symmetric ones. On finite-dimensional complex semisimple Lie algebras and finite-dimensional modules, they vanish identically. Liu, Liu, and Zhao proved that if \(\mathfrak{g}\) is finite-dimensional complex semisimple and \(V\) is a finite-dimensional \(\mathfrak{g}\)-module, then every symmetric biderivation
\[
\delta:\mathfrak{g}\times \mathfrak{g}\to V
\]
is trivial [2407.05581]. In particular, symmetric adjoint-valued biderivations on semisimple Lie algebras are zero [2407.05581].

The proof uses Whitehead’s lemma, complete reducibility, and systematic reduction to \(\mathfrak{sl}_2\)-subalgebras. A key device is a proposition asserting that if \(\delta:\mathfrak{g}\times V\to W\) is a bilinear map that is a module homomorphism in one slot and a derivation in the other, then \(\delta\) must vanish when \(\mathfrak{g}\) is semisimple and \(V\) is finite-dimensional [2407.05581]. This mechanism underlies the vanishing theorem and its applications to Takiff algebras, symplectic oscillator algebras, Schrödinger algebras, and certain Lie superalgebras [2407.05581].

This semisimple vanishing result was sharpened in the perfect case. In 2025, it was shown that there are no nonzero symmetric biderivations on finite-dimensional perfect Lie algebras over a field of characteristic zero, even for values in arbitrary finite-dimensional modules [2503.13059]. Equivalently, every symmetric biderivation
\[
\phi:g\times g\to M
\]
on a finite-dimensional perfect Lie algebra \(g\) with values in a finite-dimensional \(g\)-module \(M\) is identically zero [2503.13059]. This resolves an open question posed by Brešar and Zhao [2503.13059].

The proof is formulated through **ABD-structures**: commutative products \(x\circ y\) such that each left multiplication is a derivation of the Lie bracket. Symmetric adjoint-valued biderivations are exactly ABD-structures [2503.13059]. The argument reduces via Levi decomposition to the case of an abelian radical, then rules out nontrivial products using \(\mathfrak{sl}_2\)-module analysis, Whitehead’s lemma, and the Jacobson–Morozov theorem [2503.13059].

These vanishing theorems delimit the settings in which symmetric biderivations can occur. They disappear on finite-dimensional semisimple and, more generally, perfect Lie algebras in characteristic zero [2407.05581], [2503.13059], but survive in non-perfect, solvable, or infinite-dimensional contexts, as several explicit classifications show.

## 4. Non-inner and non-skew biderivations beyond the simple case

The archetypal finite-dimensional non-simple example is \(\mathfrak{gl}_n(\mathbb{C})\). Using the decomposition
\[
\mathfrak{gl}_n=\mathbb{C}I_n\oplus \mathfrak{sl}_n,\qquad [\mathfrak{gl}_n,\mathfrak{gl}_n]=\mathfrak{sl}_n,
\]
Tang proved that a bilinear map \(D:\mathfrak{gl}_n\times \mathfrak{gl}_n\to\mathfrak{gl}_n\) is a biderivation if and only if
\[
D(X,Y)=p\,\mathrm{tr}(X)\,\mathrm{tr}(Y)\,I_n+\lambda[X,Y]
\]
for some \(\lambda,p\in\mathbb{C}\) [1610.03765]. The space of biderivations is therefore two-dimensional, spanned by the bracket and the symmetric central trace term [1610.03765]. When \(p\neq 0\), this yields explicit biderivations that are non-inner and non-skew-symmetric [1610.03765]. The phenomenon reflects the nontrivial center \(\mathbb{C}I_n\), absent in simple Lie algebras [1610.03765].

Infinite-dimensional Lie algebras exhibit a broader range of non-inner behavior. For the Schrödinger–Virasoro algebra \(SV(\varepsilon)\), Tang obtained a complete classification:
\[
f(x,y)=\lambda[x,y]+X_Q(x,y),
\]
where \(X_Q\) is supported only on \((L_m,L_n)\)-pairs and takes values in the \(M\)-sector through a finite-support family \(Q=\{\mu_k\}\) [1611.04082]. These \(X_Q\) are symmetric, non-inner, and non-skew-symmetric whenever \(Q\neq 0\) [1611.04082]. In contrast, every skew-symmetric biderivation on \(SV(\varepsilon)\) is inner [1611.04082].

A similar parameter-sensitive classification appears for the Lie algebras \(W(a,b)\). Every biderivation of \(W(a,b)\) has one of the following forms:
\[
f(x,y)=\lambda[x,y]+V_\Omega(x,y)\quad (b=0),
\]
\[
f(x,y)=\lambda[x,y]+Y_\Omega(x,y)\quad (b=1),
\]
\[
f(x,y)=\lambda[x,y]+O_a(x,y)\quad (b=-1,\ a\in \mathbb{Z}),
\]
and otherwise only the inner term survives [1712.09202]. Here \(V_\Omega\) and \(Y_\Omega\) are symmetric non-inner families, while \(O_a\) is skew-symmetric and non-inner [1712.09202]. This shows that non-inner symmetric biderivations exist precisely for \(b\in\{0,1\}\), whereas non-inner skew-symmetric ones occur in the resonant case \(b=-1\), \(a\in\mathbb{Z}\) [1712.09202].

The deformative Schrödinger–Virasoro Lie algebras \(L(\lambda,u,s)\) provide another infinite-dimensional family. In the skew-symmetric category, every biderivation is inner except when \(\lambda=1\) and \(u\in s+\mathbb{Z}\), where two explicitly constructed non-inner families \(\phi_0\) and \(\phi_1\) appear [1611.05718]. In those special regimes, the vector space of skew-symmetric biderivations has dimension two or three depending on the arithmetic condition on \(u\) [1611.05718].

These examples show a common pattern. Rigidity dominates in perfect centerless settings, but nontrivial center, solvable extensions, or infinite-dimensional grading structures create room for extra central or module-valued components. This suggests that failures of perfectness or centerlessness are the primary sources of non-inner biderivations, a conclusion made explicit in several classifications [1610.03765], [1712.09202], [1611.04082].

## 5. Commuting maps, post-Lie structures, and related constructions

Biderivations are closely linked to commuting linear maps. A linear map \(\Phi:L\to L\) is commuting if
\[
[\Phi(x),x]=0\qquad \forall x\in L.
\]
If \(\Phi\) is commuting, then
\[
D(x,y)=[\Phi(x),y]=[x,\Phi(y)]
\]
is a biderivation; conversely, biderivations of this special form recover commuting maps [1610.03765]. On finite-dimensional complex simple Lie algebras, commuting maps are exactly scalar multiples of the identity:
\[
\Phi=\lambda\,\mathrm{id}_L
\]
[1610.03765]. On \(\mathfrak{gl}_n(\mathbb{C})\), they have the form
\[
\Phi(X)=\sigma(X)I_n+\lambda X
\]
for a linear functional \(\sigma\) and a scalar \(\lambda\) [1610.03765].

Brešar and Zhao established a complementary module-valued rigidity statement: if \(L\) is a Lie algebra and \(M\) an \(L\)-module with \(Z_M(L')=0\), every commuting linear map \(f:L\to M\) lies in the centroid \(\operatorname{Cent}(M)\) [1801.01109]. In the adjoint case, under the mild condition \([c,[L,L]]=0\Rightarrow c=0\), commuting maps \(L\to L\) are centroidal [1801.01109].

Symmetric biderivations also control commutative post-Lie algebra structures. A commutative post-Lie product \(x\cdot y\) is symmetric and satisfies
\[
[x,y]\cdot z=x\cdot (y\cdot z)-y\cdot (x\cdot z),\qquad
x\cdot [y,z]=[x\cdot y,z]+[y,x\cdot z].
\]
The bilinear map \((x,y)\mapsto x\cdot y\) is then a symmetric biderivation [1611.04082], [1712.09202], [2406.06376]. As a consequence, vanishing theorems for symmetric biderivations imply triviality of commutative post-Lie structures in many rigid settings. This yields triviality on finite-dimensional complex semisimple Lie algebras [2407.05581], on finite-dimensional perfect Lie algebras in characteristic zero via ABD-structure nonexistence [2503.13059], on Schrödinger–Virasoro algebras [1611.04082], on \(W(a,b)\) [1712.09202], and on affine-Virasoro Lie algebras [2508.15573].

The affine-Virasoro case is especially strong. For the affine-Virasoro Lie algebra \(L(\mathfrak{g})\) associated to a finite-dimensional complex simple \(\mathfrak{g}\), every derivation is inner, every skew-symmetric biderivation is inner, every symmetric biderivation is trivial, and hence every biderivation is inner [2508.15573]. The same rigidity then forces every commutative post-Lie algebra structure on \(L(\mathfrak{g})\) to be trivial [2508.15573].

These relationships place biderivations at the intersection of derivation theory, centroid theory, and compatible algebraic structures. In many papers, commuting maps and post-Lie structures are not separate topics but direct corollaries of biderivation classifications [1610.03765], [1712.09202], [1611.04082], [2508.15573].

## 6. The space of biderivations as an algebraic object

For classical two-sided Lie biderivations, the literature has largely treated \(\operatorname{BiDer}(L)\) as a vector space subject to classification rather than as a Lie algebra in its own right. Recent work changes this perspective by introducing Lie brackets on spaces of **right** and **left** biderivations [2507.12089].

A right biderivation is a bilinear map \(B:g\times g\to g\) that is linear in the first argument and satisfies
\[
B([x,y],z)=[x,B(y,z)]+[B(x,z),y].
\]
For \(B_1,B_2\in \operatorname{BiDer}^r(g)\), the bracket
\[
B_1\rhd B_2(x,y)=B_1(B_2(x,y),y)-B_2(B_1(x,y),y)
\]
makes \(\operatorname{BiDer}^r(g)\) into a Lie algebra [2507.12089]. Fiberwise, for fixed \(y\), this is the commutator of derivations \(B_i(\cdot,y)\in \operatorname{Der}(g)\) [2507.12089]. Dually, left biderivations carry a Lie bracket
\[
B_1\lhd B_2(x,y)=B_1(x,B_2(x,y))-B_2(x,B_1(x,y)),
\]
and the corresponding space is likewise a Lie algebra [2507.12089].

For symmetric or skew-symmetric right biderivations, left and right notions coincide; such maps are automatically two-sided [2507.12089]. The two Lie brackets are then related by transpose:
\[
B_1\rhd B_2=(B_1^t\lhd B_2^t)^t
\]
[2507.12089]. However, even when \(B_1\) and \(B_2\) are symmetric, their right bracket need not remain symmetric or left-sided, as explicit computations on the Heisenberg algebra show [2507.12089].

This development is distinct from the classical classification results. It does not claim that the traditional space of two-sided biderivations on an arbitrary Lie algebra is itself closed under a natural commutator. Rather, it constructs a Lie algebra structure on the one-sided spaces \(\operatorname{BiDer}^r(g)\) and \(\operatorname{BiDer}^l(g)\), with the two-sided space appearing as their intersection [2507.12089]. A distinguished closed subalgebra consists of maps
\[
B(x,y)=g(y)F(x)
\]
with \(F\in \operatorname{Der}(g)\); these are stable under the right bracket and can be integrated, for simply connected Lie groups, to curves of automorphisms [2507.12089].

A different but related algebraic viewpoint occurs in Leibniz theory, where a biderivation is defined as a pair \((d,D)\) of a derivation and an anti-derivation satisfying a compatibility relation, and the resulting space carries a natural right Leibniz algebra structure [2209.07892], [2507.18134]. In solvable Leibniz algebras with null-filiform or filiform nilradicals, all biderivations are inner and the biderivation algebra is isomorphic to the original algebra via \(x\mapsto(-\operatorname{ad}_x,\operatorname{Ad}_x)\) [2507.18134]. This is not a Lie-algebra result in the strict sense, but it underscores that “algebra of biderivations” can mean more than a mere vector space, depending on the ambient nonassociative category.

The recent bracket construction for right and left biderivations suggests an emerging shift from classification alone to internal algebraic structure [2507.12089]. A plausible implication is that biderivations may eventually play a role analogous to ordinary derivations in deformation theory and higher-order symmetry analysis, although the current literature mostly develops the formal bracket and its immediate consequences rather than a full cohomological framework [2507.12089].

## 7. Scope, limitations, and current picture

The current theory is sharply stratified by structural hypotheses.

For finite-dimensional complex simple Lie algebras, the classification is complete: every biderivation is \(\lambda[\cdot,\cdot]\), symmetric biderivations vanish, commuting maps are scalar, and associated commutative post-Lie structures are trivial [1610.03765], [2407.05581]. For finite-dimensional semisimple Lie algebras, the space of biderivations is identified with the centroid, one scalar per simple summand [2307.15750]. For finite-dimensional perfect Lie algebras in characteristic zero, symmetric biderivations vanish even with module values [2503.13059].

Beyond that regime, extra phenomena appear systematically. A nontrivial center yields additional symmetric central terms, as in \(\mathfrak{gl}_n\) [1610.03765]. Infinite-dimensional graded or semidirect-product Lie algebras may admit large families of non-inner symmetric or skew-symmetric biderivations, typically concentrated in abelian or central layers, as in Schrödinger–Virasoro and \(W(a,b)\) [1611.04082], [1712.09202]. These examples show that “all biderivations are inner” is not a generic statement but a rigidity phenomenon tied to simplicity, perfectness, and centerlessness.

Characteristic assumptions are essential. Several vanishing and classification theorems rely on Whitehead’s lemma, Levi decomposition, Jacobson–Morozov theory, or root-space arguments requiring characteristic zero or exclusions such as \(\operatorname{char}(F)\neq 2,3\) [2503.13059], [2406.06376]. The existing results do not claim parallel classifications in positive characteristic, and several proofs are explicitly tied to characteristic-zero representation theory [2503.13059].

Finally, the notion of “Lie algebra of biderivations” is still not uniform across the literature. In the classical Lie-algebra setting it often means the classified vector space \(\operatorname{BiDer}(L)\); in recent work on one-sided biderivations it means an actual Lie algebra under a newly defined bracket [2507.12089]; and in Leibniz settings it can mean a right Leibniz algebra of compatible derivation–anti-derivation pairs [2209.07892], [2507.18134]. The common core across these variants is the same: biderivations encode second-order derivation behavior, and their structure is governed by the tension between rigidity from inner derivations and flexibility from centers, radicals, and grading-induced module effects.

Source: https://www.emergentmind.com/topics/lie-algebra-of-biderivations