Classification of the Real Structures of the Operators of the Biquaternion Algebra, including Lie Algebras and Lie Groups in $M_2(\mathbb{C})$

Introduction

The question. The complex operator algebra is $\mathcal{E}_{\mathbb{C}}=\operatorname{End}_{\mathbb{C}}(M_2(\mathbb{C}))\cong M_4(\mathbb{C})$, with the two real forms $M_4(\mathbb{R})$ and $M_2(\mathbb{H})$. This article describes the two forms layer by layer: the associative subalgebras, the Lie subalgebras with their Killing forms, the Jordan layer, and the Lie groups attached to the Lie algebras.

Why it is the same statement. This is the in-depth matrix reading of Classification of the Real Structures of the Operators of the Biquaternion Algebra, including Lie Algebras and Lie Groups. The source works in the abstract operator algebra; the display identifies it with $M_4(\mathbb{C})$ and writes every subalgebra as a set of explicit matrices. The statements are the same statements, and only the writing is new.

What the article proves. In this order. The two real forms, with the trace-form inertia that separates them. The biquaternion base inside each form: the regular families $\mathrm I_2\otimes M_2(\mathbb{C})$ and $M_2(\mathbb{C})^{\mathsf T}\otimes\mathrm I_2$, and the quaternion copy $L(\mathbb{H})$. The module dictionary for the semisimple associative subalgebras of $M_4(\mathbb{R})$, $\sum_i m_in_id_i=4$, and the seventeen isomorphism types. The corresponding dictionary for $M_2(\mathbb{H})$, $\sum_in_i\le2$. The Lie layer: $\mathfrak{gl}(4,\mathbb{R})=\mathfrak{sl}(4,\mathbb{R})\oplus\mathbb{R}\mathrm I_4$, the Killing form $8\operatorname{tr}(XY)-2\operatorname{tr}X\operatorname{tr}Y$, the inertia table, and the maximal subalgebras. The quaternionic Lie layer, with $\mathfrak{sl}(2,\mathbb{H})\cong\mathfrak{su}^{*}(4)$. The Jordan layer. The Lie groups, with the simply connected covers and the fundamental groups.

What the article does not do. It does not reprove the classification of the real forms, which is Classification of the Structures of the Operators of the Biquaternion Algebra in $M_2(\mathbb{C})$, and it does not treat the complex structures of the split extension, which is Classification of the Complex Structures of the Operators of the Biquaternion Algebra, including Lie Algebras and Lie Groups in $M_2(\mathbb{C})$. It states the Lie correspondence and does not prove it.

Conventions. The display is $\mathsf M_2(\tilde Q)=Q_t\mathrm I_2-i(Q_x\sigma_1+Q_y\sigma_2+Q_z\sigma_3)$. The operator frame is the basis $(E_t,E_x,E_y,E_z)$ with $E_t=\mathrm I_2$, $E_x=-i\sigma_1$, $E_y=-i\sigma_2$, $E_z=-i\sigma_3$, and operators are $4\times4$ matrices in that basis. The left and right actors are $\mathrm I_2\otimes A$ and $B^{\mathsf T}\otimes\mathrm I_2$. The inertia of a symmetric form is the pair (positive, negative) of eigenvalue counts, so a compact simple algebra reads $(0,\dim)$.

Boundaries. The cone of products is The Two-Sided Operators and the Segre Cone at $n=2$. The polar decomposition of an operator is The Adjoints and the Polar Decomposition of an Operator at $n=2$. The spin groups are The Spin Group via $4\times4$ Real Matrices. The $n=2$ spine is The One-Sided Operators at $n=2$ and The Symmetric and Antisymmetric Parts of the Operator Space at $n=2$, and the display is $M_2(\mathbb{C})$ Pauli Representation : Operators.

Letters. $\mathcal{E}_{\mathbb{H}}$ is $M_2(\mathbb{H})$ in the block model. $\mathfrak{sl}_2$ is the traceless part of $M_2(\mathbb{C})$. $K$ is the Killing form. $\mathcal{E}(A_0)$ is the two-sided span of a real form $A_0$.

The Two Real Forms

The algebra and its forms. $\mathcal{E}_{\mathbb{C}}=M_4(\mathbb{C})$ has complex dimension sixteen and real dimension thirty-two. Its two real forms are $$ M_4(\mathbb{R})\quad(\text{split}),\qquad M_2(\mathbb{H})\quad(\text{quaternionic}), $$ both of real dimension sixteen, both central simple over $\mathbb{R}$.

The two real structures. $\tau_{\mathbb{R}}(X)=\bar X$ has fixed set $M_4(\mathbb{R})$, and $\tau_{\mathbb{H}}(X)=J_0\bar XJ_0^{-1}$ has fixed set the block model $$ \mathcal{E}_{\mathbb{H}}=\left\{\begin{pmatrix}A&B\\-\bar B&\bar A\end{pmatrix}:A,B\in M_2(\mathbb{C})\right\}. $$

The table.

form $\dim_{\mathbb{R}}$ centre trace-form inertia simple part Lie algebra
$M_4(\mathbb{R})$ $16$ $\mathbb{R}\mathrm I_4$ $(10,6)$ $\mathfrak{sl}(4,\mathbb{R})$ $\mathfrak{gl}(4,\mathbb{R})$
$M_2(\mathbb{H})$ $16$ $\mathbb{R}\mathrm I_4$ $(6,10)$ $\mathfrak{sl}(2,\mathbb{H})=\mathfrak{su}^{*}(4)$ $\mathfrak{gl}(2,\mathbb{H})$

Why not compact. The compact real form of $\mathfrak{gl}(4,\mathbb{C})$ is $\mathfrak{u}(4)$, and no real form of the associative algebra has that Lie algebra, since the product of two skew-Hermitian matrices is skew-Hermitian only when the two commute. The simple part of the quaternionic form has maximal compact subalgebra of dimension ten out of fifteen. The words are split and quaternionic.

The Biquaternion Base Inside Each Form

The two regular families. $\mathrm I_2\otimes M_2(\mathbb{C})$ and $M_2(\mathbb{C})^{\mathsf T}\otimes\mathrm I_2$ are commuting copies of $M_2(\mathbb{C})$, meeting in $\mathbb{C}\mathrm I_4$. Both consist of complex matrices, so both lie in the complexification of either real form; their real points are described by the next rows.

In the split form. The real points of the left family are $\mathrm I_2\otimes M_2(\mathbb{R})$, of real dimension four, and those of the right family are $M_2(\mathbb{R})^{\mathsf T}\otimes\mathrm I_2$. The two-sided span of $M_2(\mathbb{R})$ is all of $M_4(\mathbb{R})$, by the tensor identity $M_2(\mathbb{R})\otimes_{\mathbb{R}}M_2(\mathbb{R})^{\mathrm{op}}\cong M_4(\mathbb{R})$.

In the quaternionic form. The real points of the left family are $$ \mathrm I_2\otimes\mathbb{H}=\left\{\begin{pmatrix}A&0\\0&A\end{pmatrix}\right\}\subset M_2(\mathbb{H}), $$ a copy of $\mathbb{H}$ inside the block model, of real dimension four. It is the image $L(\mathbb{B}_H)$ of the quaternion space, and its traceless part is $\mathfrak{su}(2)$.

The three distinguished base objects. $L(M_2(\mathbb{C}))\cong M_2(\mathbb{C})$ of complex dimension four, $L(\mathbb{H})\cong\mathbb{H}$ of real dimension four, and $L(\mathfrak{sl}_2)$ of complex dimension three. The first is a maximal semisimple subalgebra of $M_4(\mathbb{R})$ up to conjugacy; the second is the smallest non-commutative one; the third is a subspace and not a subalgebra.

Associative Subalgebras of the Split Form

The principle. A unital subalgebra of $M_4(\mathbb{R})$ is the same thing as a faithful module structure on $\mathbb{R}^4$, so the classification is a classification of module structures.

Theorem (module dictionary). Let $A\cong\prod_{i=1}^{r}M_{n_i}(D_i)$ be semisimple with $D_i\in\{\mathbb{R},\mathbb{C},\mathbb{H}\}$ and $d_i=\dim_{\mathbb{R}}D_i$. Then $A$ embeds unitally in $M_4(\mathbb{R})$ if and only if there are integers $m_i\ge1$ with $$ \sum_{i=1}^{r}m_in_id_i=4 . $$

Proof. The simple module of $M_{n_i}(D_i)$ is $D_i^{n_i}$ of real dimension $n_id_i$, and faithfulness forces $m_i\ge1$ on each factor.

The count. Exactly seventeen isomorphism types, with weights $1$ for $\mathbb{R}$, $2$ for $M_2(\mathbb{R})$ and $\mathbb{C}$, $3$ for $M_3(\mathbb{R})$, and $4$ for $M_4(\mathbb{R})$, $M_2(\mathbb{C})$ and $\mathbb{H}$. The largest proper type is $M_3(\mathbb{R})\oplus\mathbb{R}$ of dimension ten; $M_3(\mathbb{R})$ alone does not embed, since $4$ is not a multiple of $3$.

The maximal semisimple types. $M_2(\mathbb{C})$ and $M_2(\mathbb{R})\oplus M_2(\mathbb{R})$, both of dimension eight. An irreducible semisimple subalgebra has one factor and $m=1$ with $nd=4$, giving $M_4(\mathbb{R})$, $M_2(\mathbb{C})$ and $\mathbb{H}$, and $\mathbb{H}$ sits inside $M_2(\mathbb{C})$.

The parabolics and the Borel. The stabilizer of a line in $\mathbb{R}^4$ has dimension $1+3+9=13$, of a $2$-plane $4+4+4=12$, of a hyperplane $13$. The stabilizer of a complete flag is the upper triangular algebra, of dimension $4+3+2+1=10$.

The biquaternion entries. $L(M_2(\mathbb{R}))$ of dimension four, $L(\mathbb{H})$ of dimension four, and $L(\mathfrak{sl}_2)$ of dimension three as a space. The first two are subalgebras; the third is not, since the product of two traceless matrices is not traceless. Witness: $E_x^{2}=-\mathrm I_2$.

Associative Subalgebras of the Quaternionic Form

The criterion. An embedding $A\hookrightarrow\operatorname{End}_{\mathbb{H}}(\mathbb{H}^2)$ is a faithful module structure over $A\otimes_{\mathbb{R}}\mathbb{H}^{\mathrm{op}}$ on a real space of dimension eight.

Theorem (module dictionary). For semisimple $A=\prod_iM_{n_i}(D_i)$ the embedding into $M_2(\mathbb{H})$ exists if and only if $$ \sum_{i=1}^{r}n_i\le2 . $$

Proof. $M_n(\mathbb{R})\otimes\mathbb{H}\cong M_n(\mathbb{H})$, $M_n(\mathbb{C})\otimes\mathbb{H}\cong M_{2n}(\mathbb{C})$ and $M_n(\mathbb{H})\otimes\mathbb{H}\cong M_{4n}(\mathbb{R})$, so each factor needs $4n_i$ real dimensions and the total is $4\sum_in_i\le8$.

The comparison. The split form requires $\sum m_in_id_i=4$ with all $m_i\ge1$; the quaternionic form requires $\sum_in_i\le2$. Neither list contains the other.

The consequences. No copy of $M_2(\mathbb{R})$ or $M_3(\mathbb{R})$ embeds in $M_2(\mathbb{H})$; the largest semisimple subalgebras are $\mathbb{H}$, $\mathbb{C}$ and $M_2(\mathbb{R})$, of real dimension four, for $\sum n_i=1$, and $\mathbb{C}\oplus\mathbb{C}$ and $\mathbb{C}\oplus\mathbb{H}$ are excluded because $\sum n_i=3$.

Lie Subalgebras of the Split Form

The algebra. The commutator turns $M_4(\mathbb{R})$ into $\mathfrak{gl}(4,\mathbb{R})$, of real dimension sixteen, with centre $\mathbb{R}\mathrm I_4$ and $$ \mathfrak{gl}(4,\mathbb{R})=\mathfrak{sl}(4,\mathbb{R})\oplus\mathbb{R}\mathrm I_4,\qquad \dim\mathfrak{sl}(4,\mathbb{R})=15, $$ and $\mathfrak{sl}(4,\mathbb{R})$ simple of type $A_3$.

The Killing form. Computed from the adjoint representation, $$ K(X,Y)=8\operatorname{tr}(XY)-2\operatorname{tr}(X)\operatorname{tr}(Y), $$ verified on random pairs to $4\cdot10^{-15}$. Its radical is the centre, and on the traceless part it is $8\operatorname{tr}(XY)$, of inertia $(9,6)$.

The inertia table. The pair is (positive, negative); a null direction from the centre is recorded separately.

subalgebra $\dim_{\mathbb{R}}$ type $K$ inertia
$\mathfrak{gl}(4,\mathbb{R})$ $16$ $A_3\oplus T_1$ $(9,6)$ plus one null
$\mathfrak{sl}(4,\mathbb{R})$ $15$ $A_3$ split $(9,6)$
$\mathfrak{sp}(2,\mathbb{R})$ $10$ $C_2$ split $(6,4)$
$\mathfrak{sl}(3,\mathbb{R})\oplus\mathbb{R}$ $9$ Levi $(5,3)$ plus one null
$\mathfrak{so}(4)$ $6$ $D_2$ compact $(0,6)$
$\mathfrak{so}(3,1)$ $6$ $D_2$ Lorentzian $(3,3)$
$\mathfrak{so}(2,2)$ $6$ $D_2$ split $(4,2)$
$\mathfrak{gl}(2,\mathbb{R})$ $4$ $A_1\oplus T_1$ $(2,1)$ plus one null
$\mathfrak{u}(2)$ $4$ compact reductive $(0,3)$ plus one null
$\mathfrak{sl}(2,\mathbb{R})$ $3$ $A_1$ split $(2,1)$
$\mathfrak{su}(2)$ $3$ $A_1$ compact $(0,3)$

The three six-dimensional algebras. $\mathfrak{so}(4)$, $\mathfrak{so}(3,1)$ and $\mathfrak{so}(2,2)$ are real forms of the one complex algebra $\mathfrak{sl}(2,\mathbb{C})\oplus\mathfrak{sl}(2,\mathbb{C})$ and are pairwise non-isomorphic, separated by their inertias.

The maximal subalgebras. Up to conjugacy: the two parabolics of dimensions twelve and eleven, the symplectic $\mathfrak{sp}(2,\mathbb{R})$ of dimension ten, the two Levi factors of dimensions nine and seven, the three $\mathfrak{so}$ algebras of dimension six, and an $\mathfrak{sl}(2,\mathbb{R})$ of dimension three acting irreducibly on $\operatorname{Sym}^3$.

The biquaternion subalgebras. $\mathfrak{sl}(2,\mathbb{R})\oplus\mathfrak{sl}(2,\mathbb{R})$ of dimension six sits inside $\mathfrak{so}(2,2)$; the compact $\mathfrak{su}(2)\oplus\mathfrak{su}(2)$ sits inside $\mathfrak{so}(4)$; and the copy $\mathfrak{u}(2)$ comes from $\mathbb{B}_{\Phi}$.

Lie Subalgebras of the Quaternionic Form

The algebra. $\mathfrak{gl}(2,\mathbb{H})$ of real dimension sixteen, with centre $\mathbb{R}\mathrm I_4$ and simple part $$ \mathfrak{sl}(2,\mathbb{H})=\mathfrak{su}^{*}(4),\qquad \dim_{\mathbb{R}}=15, $$ whose maximal compact subalgebra is $\mathfrak{sp}(2)$ of dimension ten and whose complement has dimension five.

The table.

subalgebra $\dim_{\mathbb{R}}$ type compact part
$\mathfrak{gl}(2,\mathbb{H})$ $16$ $A_3$ quaternionic $+T_1$ $\mathfrak{sp}(2)\oplus\mathbb{R}$
$\mathfrak{sl}(2,\mathbb{H})=\mathfrak{su}^{*}(4)$ $15$ $A_3$ quaternionic $\mathfrak{sp}(2)$, $\dim10$
$\mathfrak{sp}(2)$ $10$ $C_2$ compact itself
$\mathfrak{sp}(1,1)$ $10$ $C_2$ non-compact $\mathfrak{sp}(1)\oplus\mathfrak{sp}(1)$
$\mathfrak{sp}(1)\oplus\mathfrak{sp}(1)\oplus\mathbb{R}$ $7$ Levi itself

The two algebras of dimension ten. $\mathfrak{sp}(2)$ and $\mathfrak{sp}(1,1)$ are real forms of $\mathfrak{sp}(2,\mathbb{C})$; the first is compact with killing inertia $(0,10)$, the second has inertia $(5,5)$.

The biquaternion copy. $\mathfrak{su}(2)=\mathfrak{sp}(1)$ sits inside $\mathfrak{sl}(2,\mathbb{H})$ as the traceless part of $L(\mathbb{H})$, and the pair $\mathfrak{sp}(1)\oplus\mathfrak{sp}(1)$ is the image of $\mathfrak{sl}_2\oplus\mathfrak{sl}_2$ under the two regular families.

The Jordan Layer

The symmetrized product. $X\circ Y=\tfrac12(XY+YX)$ makes $M_4(\mathbb{C})^{+}$ a Jordan algebra, with the three real forms $$ M_4(\mathbb{R})^{+},\qquad M_2(\mathbb{H})^{+},\qquad H_4(\mathbb{C}), $$ of real dimension sixteen and trace-form inertias $(10,6)$, $(6,10)$ and $(16,0)$.

The Peirce layer. The idempotents of $M_4(\mathbb{C})^{+}$ include the two diagonal blocks $P_1\oplus P_2$ and the four coordinate projectors, and the Peirce decomposition of a Jordan algebra with two orthogonal idempotents is the matrix form of the corner decomposition of One-Sided Operators on the N-Regular Sector in $M_2(\mathbb{C})$.

The Hermitian subspace. $H_4(\mathbb{C})$ is closed under $\circ$, is not closed under the associative product, and has no associative real form above it. Its Lie shadow is $\mathfrak{u}(4)$, the compact form that has no associative model.

The Lie Groups Attached to the Lie Algebras

The correspondence. Each real Lie algebra has a simply connected group, unique up to isomorphism; every connected group with that algebra is a quotient by a discrete central subgroup, and the fundamental group of the quotient is that subgroup.

The split form.

algebra simply connected group centre noteworthy quotient
$\mathfrak{gl}(4,\mathbb{R})$ $\widetilde{SL(4,\mathbb{R})}\times\mathbb{R}$ $\mathbb{Z}_2^{2}\times\mathbb{R}$ $GL^{+}(4,\mathbb{R})\cong SL(4,\mathbb{R})\times\mathbb{R}_{>0}$, $\pi_1=\mathbb{Z}_2$
$\mathfrak{sl}(4,\mathbb{R})$ $\widetilde{SL(4,\mathbb{R})}$ $\mathbb{Z}_2^{2}$ $SL(4,\mathbb{R})$, $\pi_1=\mathbb{Z}_2$; $PSL(4,\mathbb{R})$, $\pi_1=\mathbb{Z}_2^{2}$
$\mathfrak{sp}(2,\mathbb{R})$ $\widetilde{Sp(2,\mathbb{R})}$ $\mathbb{Z}$ $Sp(2,\mathbb{R})$, $\pi_1=\mathbb{Z}$
$\mathfrak{so}(4)$ $Spin(4)=SU(2)\times SU(2)$ $\mathbb{Z}_2^{2}$ $SO(4)$, $\pi_1=\mathbb{Z}_2$
$\mathfrak{so}(3,1)$ $SL(2,\mathbb{C})$ $\mathbb{Z}_2$ $SO^{+}(3,1)$, $\pi_1=\mathbb{Z}_2$
$\mathfrak{so}(2,2)$ $\widetilde{SL(2,\mathbb{R})}^{2}$ $\mathbb{Z}^{2}$ $SO^{+}(2,2)\cong(SL(2,\mathbb{R})\times SL(2,\mathbb{R}))/\{\pm(\mathrm I,\mathrm I)\}$
$\mathfrak{su}(2)$ $SU(2)$ $\mathbb{Z}_2$ $SO(3)$, $\pi_1=\mathbb{Z}_2$

The quaternionic form.

algebra simply connected group centre noteworthy quotient
$\mathfrak{gl}(2,\mathbb{H})$ $SL(2,\mathbb{H})\times\mathbb{R}$ $\mathbb{Z}_2\times\mathbb{R}$ $GL(2,\mathbb{H})\cong SL(2,\mathbb{H})\times\mathbb{R}_{>0}$, $\pi_1=0$
$\mathfrak{sl}(2,\mathbb{H})$ $SL(2,\mathbb{H})$ $\mathbb{Z}_2$ $PSL(2,\mathbb{H})\cong SO^{+}(5,1)$, $\pi_1=\mathbb{Z}_2$
$\mathfrak{sp}(2)$ $Sp(2)$ $\mathbb{Z}_2$ $SO(5)$, $\pi_1=\mathbb{Z}_2$
$\mathfrak{sp}(1,1)$ $Spin(4,1)$ $\mathbb{Z}_2$ $SO^{+}(4,1)$, $\pi_1=\mathbb{Z}_2$

The five groups of $\mathfrak{sl}(4,\mathbb{R})$. The centre of the cover is $\mathbb{Z}_2^{2}$, whose five subgroups give five connected groups: the cover, $SL(4,\mathbb{R})$, $PSL(4,\mathbb{R})$ and two intermediate double covers. The element level has three groups for $\mathfrak{sl}(2,\mathbb{R})$, and the difference is the rank of the centre.

The simply connected quaternionic groups. $SL(2,\mathbb{H})$ and $GL(2,\mathbb{H})$ are simply connected, because their maximal compact subgroup is $Sp(2)$, which is simply connected. The real counterparts have double covers because $SO(4)$ does not.

The three covers used in geometry. $$ SL(2,\mathbb{C})\to SO^{+}(3,1),\qquad Spin(4)=SU(2)\times SU(2)\to SO(4),\qquad SL(2,\mathbb{H})\to SO^{+}(5,1). $$

The algebra does not determine the group. $\mathfrak{sl}(4,\mathbb{R})$ is the algebra of five connected groups and $\mathfrak{gl}(4,\mathbb{R})$ of infinitely many; the discrete central subgroup is extra data.

Where $n=2$ Differs

The real rank. The simple part $\mathfrak{sl}(4,\mathbb{R})$ has real rank three against the real rank one of $\mathfrak{sl}(2,\mathbb{R})$ at the element level. The three commuting $\mathfrak{sl}(2,\mathbb{R})$'s of the split form are the origin of the three rotation planes of the biquaternion algebra, and at $n=2$ there is one.

The dimension of the complement. $\mathfrak{sl}_2\otimes\mathfrak{sl}_2$ is nine-dimensional at $n=2$, that is $3\times3$; the Cartan complement of the maximal compact is five-dimensional in the quaternionic form and nine-dimensional in the split one, against one at the element level.

The group count. $\mathfrak{sl}(4,\mathbb{R})$ has five connected groups and $\mathfrak{sl}(2,\mathbb{R})$ three; the counts differ because the centre of the cover is $\mathbb{Z}_2^{2}$ against $\mathbb{Z}$.

Nothing else differs in kind. The module dictionaries, the maximal subalgebras, the quaternionic simple part $\mathfrak{su}^{*}(4)$ and the Jordan triple hold at every order with the sizes changed.

Summary

The operator algebra $M_4(\mathbb{C})$ has the two real forms $M_4(\mathbb{R})$ and $M_2(\mathbb{H})$, separated by the trace-form inertia $(10,6)$ against $(6,10)$. Inside each form sits the biquaternion base: the regular families $\mathrm I_2\otimes M_2(\mathbb{R})$ and $M_2(\mathbb{R})^{\mathsf T}\otimes\mathrm I_2$ in the split form, and the copies $\mathrm I_2\otimes\mathbb{H}$ and $\mathbb{H}\otimes\mathrm I_2$ in the quaternionic one.

The semisimple associative subalgebras are classified by the module dictionaries $\sum m_in_id_i=4$ for the split form, with seventeen isomorphism types and the maximal pair $M_2(\mathbb{C})$, $M_2(\mathbb{R})\oplus M_2(\mathbb{R})$, and $\sum_in_i\le2$ for the quaternionic form.

The Lie layer of the split form is $\mathfrak{gl}(4,\mathbb{R})=\mathfrak{sl}(4,\mathbb{R})\oplus\mathbb{R}\mathrm I_4$ with Killing form $8\operatorname{tr}(XY)-2\operatorname{tr}X\operatorname{tr}Y$, verified to $4\cdot10^{-15}$, and inertia $(9,6)$ on the traceless part. The quaternionic simple part is $\mathfrak{su}^{*}(4)$ with compact $\mathfrak{sp}(2)$ of dimension ten.

The Jordan layer has the three real forms $M_4(\mathbb{R})^{+}$, $M_2(\mathbb{H})^{+}$ and $H_4(\mathbb{C})$. The groups attached to the split algebra are quotients of $\widetilde{SL(4,\mathbb{R})}$ by the five subgroups of $\mathbb{Z}_2^{2}$, and the quaternionic groups $SL(2,\mathbb{H})$ and $GL(2,\mathbb{H})$ are simply connected, with $SL(2,\mathbb{H})\cong Spin(5,1)$.

Summary of Notation

Symbol Meaning
$\mathcal{E}_{\mathbb{H}}$ $M_2(\mathbb{H})$ in the block model
$m_i,n_i,d_i$ multiplicity, matrix size and division-algebra dimension
$K$ the Killing form
$\mathfrak{su}^{*}(4)$ $\mathfrak{sl}(2,\mathbb{H})$
$\mathfrak{sp}(2),\mathfrak{sp}(1,1)$ the compact and the non-compact form of $C_2$
$H_4(\mathbb{C})$ the Hermitian matrices

Further Reading