The Three Canonical Forms on $\mathbb{C}^2$
Introduction
The duo-complex numbers carry three canonical forms: the Hermitian form $H$, the complex bilinear symmetric form $B$ and the complex bilinear alternating form $\omega$. They are defined in Duo-Complex Numbers and announced there; this article computes them. It gives their Gram matrices, their symmetry and nondegeneracy, their isotropic sets, their real and imaginary parts, the relations among them, and the distinction between $\operatorname{Im}H$ and $\omega$, which is the point at which the three are most often confused.
The article treats the forms on the space alone. The isometry groups are The Classical Groups of $\mathbb{C}^2$; the Kähler structure they define is The Flat Kähler Structure of $\mathbb{C}^2$; the real inertias are also read in The Realification of $\mathbb{C}^2$, and the computations are repeated here only where the complex statement needs them. The general theory of bilinear and sesquilinear forms is Bilinear Forms and The Sesquilinear Structure of the Endomorphism Algebra; this article does not restate it.
The three forms are of three different kinds and are not three normalisations of one object. $H$ is sesquilinear and definite; $B$ is symmetric and indefinite; $\omega$ is alternating and nondegenerate. The first carries the metric, the second the isotropic cone, the third the holomorphic symplectic structure. One table at the end of the article collects them. Their use does not stop at one system: carried to tensor powers the three become the metric, the norm and the entanglement pairing, which is the subject of The Duo-Complex Bridge: The Three Canonical Forms and Their Tensor Powers. There the tensor square of $\omega$ evaluates on a two-qubit state to $2\det c$, the determinant of the coefficient array, so the concurrence of two qubits is the alternating form of this article applied to both factors.
The Hermitian Form
The Hermitian form is the positive definite sesquilinear form of the space; its Gram matrix is the identity and its norm is the Euclidean norm.
Definition. The Hermitian form is $$ H(Q_1,Q_2)=Q_{1\blacktriangle}\overline{Q_{2\blacktriangle}}+Q_{1\blacktriangledown}\overline{Q_{2\blacktriangledown}}. $$
Proposition (the Gram matrix). In the complex basis $e_1,e_2$ the Gram matrix is the identity, $$ \bigl(H(e_j,e_k)\bigr)_{j,k}=\begin{pmatrix}1&0\\0&1\end{pmatrix}. $$
Proof. $H(e_1,e_1)=1$, $H(e_1,e_2)=0$, $H(e_2,e_1)=0$, $H(e_2,e_2)=1$.
Proposition (sesquilinearity and positivity). The form is linear in the first slot and conjugate-linear in the second, $$ H(\lambda Q_1,Q_2)=\lambda H(Q_1,Q_2),\qquad H(Q_1,\lambda Q_2)=\overline{\lambda}H(Q_1,Q_2), $$ it is Hermitian, $H(Q_2,Q_1)=\overline{H(Q_1,Q_2)}$, and it is positive definite, $H(Q,Q)>0$ for $Q\neq0$.
Proof. The two linearity rules are read on the definition. The Hermitian symmetry is the identity $\overline{Q_{1\blacktriangle}\overline{Q_{2\blacktriangle}}}=Q_{2\blacktriangle}\overline{Q_{1\blacktriangle}}$, and likewise for the descending triangle. The positivity is $\lvert Q_{\blacktriangle}\rvert^2+\lvert Q_{\blacktriangledown}\rvert^2$, a sum of non-negative terms, zero only when both coordinates vanish.
Remark (the only definite form of the three). The form $H$ is the only one of the three whose isotropic set is the origin. Its inertia over $\mathbb{R}$ is $(4,0)$, and it is the form that defines the topology of the space. The other two have non-trivial isotropic sets and indefinite real pictures; their role is geometric and not metric.
The real and imaginary parts of $H$ are computed in The Realification of $\mathbb{C}^2$: the real part is the Euclidean metric $g$ of inertia $(4,0)$ and the imaginary part is the symplectic form $-\Omega$, where $\Omega=dq_{\blacktriangle}\wedge dq'_{\blacktriangle}+dq_{\blacktriangledown}\wedge dq'_{\blacktriangledown}$ is the Kähler form. The relation is $$ H(Q_1,Q_2)=g(Q_1,Q_2)-i\Omega(Q_1,Q_2),\qquad \Omega(Q_1,Q_2)=g(JQ_1,Q_2), $$ and it is the reason the Hermitian form carries both the metric and the symplectic structure of the space.
The Complex Bilinear Form $B$
The complex bilinear form is symmetric and nondegenerate; its isotropic set is the union of two complex lines, and in suitable holomorphic coordinates its quadratic form is the product $uv$.
Definition. The complex bilinear form is $$ B(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangle}+Q_{1\blacktriangledown}Q_{2\blacktriangledown}. $$
Proposition (symmetry, Gram matrix and nondegeneracy). The form is $\mathbb{C}$-bilinear and symmetric, $B(Q_2,Q_1)=B(Q_1,Q_2)$; its Gram matrix in the complex basis is the identity; and it is nondegenerate.
Proof. Bilinearity and symmetry are read on the definition. The Gram matrix is $B(e_j,e_k)=\delta_{jk}$, of determinant $1$, hence invertible, which is nondegeneracy.
Theorem (the isotropic cone). The isotropic set of $B$ is the union of the two complex lines $$ L_+=\{Q:Q_{\blacktriangle}=iQ_{\blacktriangledown}\},\qquad L_-=\{Q:Q_{\blacktriangle}=-iQ_{\blacktriangledown}\}, $$ and the two lines meet only at the origin.
Proof. $B(Q,Q)=Q_{\blacktriangle}^2+Q_{\blacktriangledown}^2=(Q_{\blacktriangle}+iQ_{\blacktriangledown})(Q_{\blacktriangle}-iQ_{\blacktriangledown})$. The field $\mathbb{C}$ is a domain, so the product vanishes exactly when one factor does, that is $Q_{\blacktriangle}=-iQ_{\blacktriangledown}$ or $Q_{\blacktriangle}=iQ_{\blacktriangledown}$. A point of both lines satisfies $2Q_{\blacktriangle}=0$ and $2iQ_{\blacktriangledown}=0$, hence is the origin.
Theorem (the normal form). In the holomorphic coordinates $$ u(Q)=Q_{\blacktriangle}+iQ_{\blacktriangledown},\qquad v(Q)=Q_{\blacktriangle}-iQ_{\blacktriangledown}, $$ the form is $$ B(Q_1,Q_2)=\tfrac12\bigl(u(Q_1)v(Q_2)+u(Q_2)v(Q_1)\bigr), $$ its quadratic form is $B(Q,Q)=u(Q)v(Q)=Q_{\blacktriangle}^2+Q_{\blacktriangledown}^2$, and the two isotropic lines are the coordinate axes $u(Q)=0$ and $v(Q)=0$.
Proof. The change of coordinates is invertible, with $Q_{\blacktriangle}=\tfrac12(u(Q)+v(Q))$ and $Q_{\blacktriangledown}=\tfrac{1}{2i}(u(Q)-v(Q))$, and $u(Q)v(Q)=(Q_{\blacktriangle}+iQ_{\blacktriangledown})(Q_{\blacktriangle}-iQ_{\blacktriangledown})=Q_{\blacktriangle}^2+Q_{\blacktriangledown}^2$. The polarization of this quadratic form is the displayed symmetric expression. The axis $u(Q)=0$ is $Q_{\blacktriangle}=-iQ_{\blacktriangledown}$, and the axis $v(Q)=0$ is $Q_{\blacktriangle}=iQ_{\blacktriangledown}$.
Remark (a hyperbolic form in complex coordinates). The normal form $B(Q_1,Q_2)=\tfrac12(u(Q_1)v(Q_2)+u(Q_2)v(Q_1))$ has the quadratic form $B(Q,Q)=u(Q)v(Q)$, the complex analogue of the hyperbolic form $x^2-y^2$ of the split-complex plane. The form is indefinite in the strongest sense: over $\mathbb{C}$ it has no inertia, because a complex number is not ordered, and its realification has inertia $(2,2)$ for both the real and the imaginary part. The absence of an inertia over $\mathbb{C}$ is not a defect of the computation: it is the statement that the form has isotropic vectors in abundance, and a definite form has none.
The two isotropic lines are complex lines, hence $J$-invariant, but they are selected by $B$ and not by $J$: every complex line is $J$-invariant, and only two of them are isotropic. They are the two points of $\mathbb{CP}^1$ at which the quadric $B=0$ meets the line at infinity (The Isotropic Cone of the Bilinear Form on $\mathbb{C}^2$).
Proposition (the behaviour under $J$). The form is anti-invariant under the complex structure, $$ B(JQ_1,JQ_2)=-B(Q_1,Q_2). $$
Proof. $B(iQ_1,iQ_2)=i^2B(Q_1,Q_2)=-B(Q_1,Q_2)$.
Remark (two different isometry groups). The unitary group preserves $H$, and the complex orthogonal group $\mathrm{O}(2,\mathbb{C})$ preserves $B$. The two groups are different, because $U^{*}U=I$ does not imply $U^{T}U=I$: a unitary matrix need not be complex orthogonal. Their intersection is the real orthogonal group $\mathrm{O}(2)$, the matrices that are at once real and unitary (The Classical Groups of $\mathbb{C}^2$).
The Holomorphic Symplectic Form $\omega$
The holomorphic symplectic form is alternating and nondegenerate; it vanishes on every complex line, and it is the form whose powers give the complex volume.
Definition. The holomorphic symplectic form is $$ \omega(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangledown}-Q_{1\blacktriangledown}Q_{2\blacktriangle}. $$
Proposition (alternation, Gram matrix and nondegeneracy). The form is $\mathbb{C}$-bilinear and alternating, $\omega(Q_2,Q_1)=-\omega(Q_1,Q_2)$; its Gram matrix in the complex basis is $$ \bigl(\omega(e_j,e_k)\bigr)_{j,k}=\begin{pmatrix}0&1\\-1&0\end{pmatrix}, $$ of determinant $1$; and it is nondegenerate.
Proof. The alternation is the exchange of the two coordinates. The entries are $\omega(e_1,e_2)=1$, $\omega(e_2,e_1)=-1$ and the two diagonal entries $\omega(e_k,e_k)=0$. The determinant is $1$.
Proposition (the form is decomposable). As a complex $2$-form, $\omega=dQ_{\blacktriangle}\wedge dQ_{\blacktriangledown}$; in particular $\omega\wedge\omega=0$ and $$ \omega\wedge\overline{\omega}=2\,\Omega^2=4\,dq_{\blacktriangle}\wedge dq'_{\blacktriangle}\wedge dq_{\blacktriangledown}\wedge dq'_{\blacktriangledown}, $$ where $\Omega=dq_{\blacktriangle}\wedge dq'_{\blacktriangle}+dq_{\blacktriangledown}\wedge dq'_{\blacktriangledown}$ is the Kähler form.
Proof. The evaluation of $dQ_{\blacktriangle}\wedge dQ_{\blacktriangledown}$ on a pair $(Q_1,Q_2)$ is $Q_{1\blacktriangle}Q_{2\blacktriangledown}-Q_{1\blacktriangledown}Q_{2\blacktriangle}$. The square of a $2$-form in two complex variables vanishes because there are only two holomorphic directions. For the last identity, $dQ_{\blacktriangle}\wedge d\overline{Q}_{\blacktriangle}=-2i\,dq_{\blacktriangle}\wedge dq'_{\blacktriangle}$ and likewise for the descending triangle, and the product of the two gives the coefficient $4$ after the sign of the reordering; the computation is one line in the real basis.
Theorem (the isotropic subspaces). For $Q_1\neq0$, $\omega(Q_1,Q_2)=0$ exactly when $Q_2$ is a complex multiple of $Q_1$. Hence the complex subspaces isotropic for $\omega$ are exactly the complex lines, and every complex line is isotropic.
Proof. If $Q_2=\lambda Q_1$, the alternation gives $\omega(Q_1,\lambda Q_1)=\lambda\omega(Q_1,Q_1)=0$. Conversely, suppose $\omega(Q_1,Q_2)=0$ with $Q_1\neq0$. If $Q_{1\blacktriangledown}\neq0$, the equation reads $Q_{2\blacktriangle}=(Q_{1\blacktriangle}/Q_{1\blacktriangledown})Q_{2\blacktriangledown}$, so $Q_2=(Q_{2\blacktriangledown}/Q_{1\blacktriangledown})\,Q_1$; the case $Q_{1\blacktriangledown}=0$ gives $Q_{1\blacktriangle}\neq0$ and $\omega(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangledown}=0$, so $Q_{2\blacktriangledown}=0$ and $Q_2=(Q_{2\blacktriangle}/Q_{1\blacktriangle})\,Q_1$.
Remark (the Lagrangian lines). A Lagrangian subspace of a symplectic space is an isotropic subspace of half the dimension. In complex dimension two the Lagrangian subspaces of $\omega$ are exactly the complex lines, and the space of them is $\mathbb{CP}^1$. This is the opposite of the behaviour of the Kähler form, on which the complex lines are symplectic and the Lagrangian planes are the totally real ones (The Subspaces of $\mathbb{C}^2$).
Proposition (the behaviour under $J$). The form is anti-invariant under the complex structure, $$ \omega(JQ_1,JQ_2)=-\omega(Q_1,Q_2), $$ so both $\operatorname{Re}\omega$ and $\operatorname{Im}\omega$ change sign under $J$.
Proof. $\omega(iQ_1,iQ_2)=i^2\omega(Q_1,Q_2)=-\omega(Q_1,Q_2)$, and the real and imaginary parts inherit the sign.
Remark (the two real symplectic forms). The real and imaginary parts of $\omega$ are $$ \operatorname{Re}\omega=dq_{\blacktriangle}\wedge dq_{\blacktriangledown}-dq'_{\blacktriangle}\wedge dq'_{\blacktriangledown},\qquad \operatorname{Im}\omega=dq_{\blacktriangle}\wedge dq'_{\blacktriangledown}-dq_{\blacktriangledown}\wedge dq'_{\blacktriangle}, $$ both nondegenerate, both anti-invariant under $J$. They are a pair of symplectic forms orthogonal under the wedge pairing, $\operatorname{Re}\omega\wedge\operatorname{Im}\omega=0$, and their squares are equal, $\operatorname{Re}\omega\wedge\operatorname{Re}\omega=\operatorname{Im}\omega\wedge\operatorname{Im}\omega$. No one of them is the Kähler form.
The Relations among the Three
The three forms are related by the conjugation of the coefficients and by the Lagrange identity; the relations are algebraic and do not make any one of them a function of the others.
Proposition (the Hermitian form is the bilinear form on the conjugate). For $Q_1,Q_2\in\mathbb{C}^2$, $$ H(Q_1,Q_2)=B(Q_1,\overline{Q}_2), $$ where $\overline{Q}_2=(\overline{Q}_{2\blacktriangle},\overline{Q}_{2\blacktriangledown})$ is the coefficientwise conjugate.
Proof. $B(Q_1,\overline{Q}_2)=Q_{1\blacktriangle}\overline{Q_{2\blacktriangle}}+Q_{1\blacktriangledown}\overline{Q_{2\blacktriangledown}}=H(Q_1,Q_2)$.
Proposition (the Lagrange identity). For all $Q_1,Q_2\in\mathbb{C}^2$, $$ B(Q_1,Q_2)^2+\omega(Q_1,Q_2)^2=B(Q_1,Q_1)\,B(Q_2,Q_2). $$
Proof. Write $a=Q_{1\blacktriangle}$, $b=Q_{1\blacktriangledown}$, $c=Q_{2\blacktriangle}$, $d=Q_{2\blacktriangledown}$. The identity is $(ac+bd)^2+(ad-bc)^2=(a^2+b^2)(c^2+d^2)$, which is the multiplicativity of the complex modulus read on the two complex numbers $a+ib$ and $c+id$.
Proposition (the normal forms in the coordinates $u,v$). Let $$ u(Q)=Q_{\blacktriangle}+iQ_{\blacktriangledown},\qquad v(Q)=Q_{\blacktriangle}-iQ_{\blacktriangledown} $$ be the two null coordinates of the point $Q$. Then $$ B(Q_1,Q_2)=\tfrac12\bigl(u(Q_1)v(Q_2)+u(Q_2)v(Q_1)\bigr),\qquad \omega(Q_1,Q_2)=\tfrac{i}{2}\bigl(u(Q_1)v(Q_2)-v(Q_1)u(Q_2)\bigr), $$ and $$ H(Q_1,Q_2)=\tfrac12\bigl(u(Q_1)\overline{u(Q_2)}+v(Q_1)\overline{v(Q_2)}\bigr). $$
Proof. The first is the theorem of the previous section. For the second, $u(Q_1)v(Q_2)-v(Q_1)u(Q_2)=-2i(Q_{1\blacktriangle}Q_{2\blacktriangledown}-Q_{1\blacktriangledown}Q_{2\blacktriangle})=-2i\,\omega(Q_1,Q_2)$. For the third, $u(Q_1)\overline{u(Q_2)}+v(Q_1)\overline{v(Q_2)}=2(Q_{1\blacktriangle}\overline{Q_{2\blacktriangle}}+Q_{1\blacktriangledown}\overline{Q_{2\blacktriangledown}})=2H(Q_1,Q_2)$.
Remark (the three forms in one coordinate system). In the coordinates $u,v$ the three forms are the product, the symplectic form and the Euclidean metric, and the space is the product of two lines. The coordinates $u,v$ are holomorphic, so the types of the forms are preserved: $B$ stays symmetric, $\omega$ stays alternating and $H$ stays Hermitian. The normal form displays the isotropic cone as the union of the axes and the Lagrangian lines as the axes themselves.
The Distinction between $\operatorname{Im}H$ and $\omega$
The forms $\operatorname{Im}H$ and $\omega$ are both alternating and both nondegenerate, and they are not the same form; the negative of the first is the Kähler form, the second is the holomorphic symplectic form.
The two are compared in the table below. The comparison is the answer to the most frequent confusion of the category.
| Property | $\operatorname{Im}H$ | $\omega$ |
|---|---|---|
| valued in | $\mathbb{R}$ | $\mathbb{C}$ |
| type as a complex form | $(1,1)$ | $(2,0)$ |
| behaviour under $J$ | invariant: $\operatorname{Im}H(JQ_1,JQ_2)=\operatorname{Im}H(Q_1,Q_2)$ | anti-invariant: $\omega(JQ_1,JQ_2)=-\omega(Q_1,Q_2)$ |
| on a complex line | minus the area form, nondegenerate | zero |
| Lagrangian subspaces | the $\Omega$-Lagrangian planes: a $3$-parameter family inside the totally real planes | the complex lines, $\mathbb{CP}^1$ |
| role | Kähler form, metric and symplectic structure | holomorphic volume form |
| relation to the metric | $\Omega=-\operatorname{Im}H=g(J\cdot,\cdot)$ | $\omega\wedge\overline{\omega}=2\Omega^2$ |
Theorem (the two are not proportional). There is no constant $\lambda\in\mathbb{C}$ with $\operatorname{Im}H=\lambda\omega$; the two forms have different values on the same pair and different isotropic subspaces.
Proof. On the pair $Q_1=(1,0)$, $Q_2=(i,0)$ one has $\operatorname{Im}H(Q_1,Q_2)=-1$ and $\omega(Q_1,Q_2)=0$. On the pair $Q_1=(1,0)$, $Q_2=(0,1)$ one has $\operatorname{Im}H(Q_1,Q_2)=0$ and $\omega(Q_1,Q_2)=1$. A proportional pair cannot take a zero and a non-zero value on the same pair, so no constant exists.
Remark (the geometric content of the distinction). The form $\operatorname{Im}H$ pairs a vector with the conjugate of another; the form $\omega$ pairs two vectors directly. The first sees the complex structure and is invariant under it; the second sees the holomorphic structure and changes sign under it. On a complex line, the first measures area and the second vanishes. This is why the Kähler form is $\Omega=-\operatorname{Im}H$ and the holomorphic symplectic form is $\omega$, and why the two are never substituted one for the other.
Remark (the consequence for the Lagrangian planes). A plane that is Lagrangian for $\Omega$ cannot contain a complex line, because $\Omega$ is nondegenerate on every complex line; hence it is totally real. The $\Omega$-Lagrangian planes are a $3$-parameter family inside the $4$-parameter family of totally real planes, and a complex line is never $\Omega$-Lagrangian. The two notions of Lagrangian therefore select different classes of planes: the $\omega$-Lagrangian lines are complex, the $\Omega$-Lagrangian planes are totally real.
Gram Matrices and the Real Pictures
The Gram matrices in the complex basis are the identity, the identity and the standard symplectic matrix; in the real basis the six real forms take the normal forms collected below.
Theorem (the complex Gram matrices). In the complex basis $e_1,e_2$, $$ G_H=\begin{pmatrix}1&0\\0&1\end{pmatrix},\qquad G_B=\begin{pmatrix}1&0\\0&1\end{pmatrix},\qquad G_\omega=\begin{pmatrix}0&1\\-1&0\end{pmatrix}. $$
Proof. The entries are computed in the three propositions above.
The real pictures are read in the real basis $e_1,ie_1,e_2,ie_2$, in this order.
| Real form | Matrix, up to the order of the blocks | Inertia or rank |
|---|---|---|
| $\operatorname{Re}H$ | $I_4$ | $(4,0)$ |
| $\operatorname{Im}H$ | $\operatorname{diag}(J_0,J_0)$ with $J_0=\begin{pmatrix}0&-1\\1&0\end{pmatrix}$ | symplectic, rank $4$ |
| $\operatorname{Re}B$ | $\operatorname{diag}(1,-1,1,-1)$ | $(2,2)$ |
| $\operatorname{Im}B$ | two blocks $\begin{pmatrix}0&1\\1&0\end{pmatrix}$ | $(2,2)$ |
| $\operatorname{Re}\omega$ | $\begin{pmatrix}0&A\\-A&0\end{pmatrix}$, $A=\operatorname{diag}(1,-1)$ | symplectic, rank $4$ |
| $\operatorname{Im}\omega$ | $\begin{pmatrix}0&B\\-B&0\end{pmatrix}$, $B=\begin{pmatrix}0&1\\1&0\end{pmatrix}$ | symplectic, rank $4$ |
Remark (the sign of the Kähler form). The form $\operatorname{Im}H$ is $-\Omega$, with $\Omega=dq_{\blacktriangle}\wedge dq'_{\blacktriangle}+dq_{\blacktriangledown}\wedge dq'_{\blacktriangledown}$. The sign is not a convention of notation but the statement that $\Omega$ is positive on the complex lines: on the line spanned by $e_1$, $\Omega(e_1,ie_1)=1$, while $\operatorname{Im}H(e_1,ie_1)=-1$. The Kähler form is taken positive, and the Hermitian form is $H=g-i\Omega$.
Summary
The space $\mathbb{C}^2$ carries three canonical forms. The Hermitian form $H(Q_1,Q_2)=Q_{1\blacktriangle}\overline{Q_{2\blacktriangle}}+Q_{1\blacktriangledown}\overline{Q_{2\blacktriangledown}}$ is sesquilinear, positive definite, of Gram matrix $I_2$ and of real inertia $(4,0)$; it gives the Euclidean norm and, with the complex structure, the decomposition $H=g-i\Omega$ into the metric and the Kähler form. The complex bilinear form $B(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangle}+Q_{1\blacktriangledown}Q_{2\blacktriangledown}$ is symmetric, nondegenerate and of Gram matrix $I_2$; its isotropic set is the union of the two complex lines $Q_{\blacktriangle}=\pm iQ_{\blacktriangledown}$, and in the holomorphic coordinates $u(Q)=Q_{\blacktriangle}+iQ_{\blacktriangledown}$, $v(Q)=Q_{\blacktriangle}-iQ_{\blacktriangledown}$ it is the product $B=uv$. The holomorphic symplectic form $\omega(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangledown}-Q_{1\blacktriangledown}Q_{2\blacktriangle}$ is alternating, nondegenerate and of Gram matrix $\begin{pmatrix}0&1\\-1&0\end{pmatrix}$; it vanishes on every complex line, so the complex lines are its Lagrangian subspaces, and its square is zero while $\omega\wedge\overline{\omega}=2\Omega^2$.
The three are related by $H(Q_1,Q_2)=B(Q_1,\overline{Q}_2)$, by the Lagrange identity $B(Q_1,Q_2)^2+\omega(Q_1,Q_2)^2=B(Q_1,Q_1)B(Q_2,Q_2)$, and by their normal forms in the coordinates $u,v$. The distinction between $\operatorname{Im}H$ and $\omega$ is the central one: $\operatorname{Im}H$ is real, of type $(1,1)$, invariant under $J$, nondegenerate on every complex line and equal to $-\Omega$; $\omega$ is complex, of type $(2,0)$, anti-invariant under $J$, zero on every complex line and related to $\Omega$ only by the quadratic identity $\omega\wedge\overline{\omega}=2\Omega^2$. The Kähler form is $\Omega=-\operatorname{Im}H$, never $\omega$.
Summary of Notation
| Symbol | Meaning |
|---|---|
| $H(Q_1,Q_2)=Q_{1\blacktriangle}\overline{Q_{2\blacktriangle}}+Q_{1\blacktriangledown}\overline{Q_{2\blacktriangledown}}$ | the Hermitian form, Gram matrix $I_2$, inertia $(4,0)$ |
| $g=\operatorname{Re}H$ | the Euclidean metric |
| $\Omega=-\operatorname{Im}H=dq_{\blacktriangle}\wedge dq'_{\blacktriangle}+dq_{\blacktriangledown}\wedge dq'_{\blacktriangledown}$ | the Kähler form, positive on the complex lines |
| $B(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangle}+Q_{1\blacktriangledown}Q_{2\blacktriangledown}$ | the complex bilinear symmetric form, Gram matrix $I_2$ |
| $L_\pm=\{Q:Q_{\blacktriangle}=\pm iQ_{\blacktriangledown}\}$ | the two isotropic lines of $B$ |
| $u(Q)=Q_{\blacktriangle}+iQ_{\blacktriangledown}$, $v(Q)=Q_{\blacktriangle}-iQ_{\blacktriangledown}$ | the holomorphic coordinates in which $B=uv$ |
| $\omega(Q_1,Q_2)=Q_{1\blacktriangle}Q_{2\blacktriangledown}-Q_{1\blacktriangledown}Q_{2\blacktriangle}=dQ_{\blacktriangle}\wedge dQ_{\blacktriangledown}$ | the holomorphic symplectic form, Gram matrix $\begin{pmatrix}0&1\\-1&0\end{pmatrix}$ |
| $\omega\wedge\overline{\omega}=2\Omega^2$ | the relation between the holomorphic and the Kähler forms |
Further Reading
- Duo-Complex Numbers, the previous article, for the definitions of the forms and the space.
- The Realification of $\mathbb{C}^2$, for the real and imaginary parts, the inertias and the dictionary.
- The Subspaces of $\mathbb{C}^2$, for the isotropic lines, the Lagrangian lines and the Lagrangian planes.
- The Isotropic Cone of the Bilinear Form on $\mathbb{C}^2$, for the cone as a real object and its image in $\mathbb{CP}^1$.
- The Flat Kähler Structure of $\mathbb{C}^2$, for the metric, the Kähler form and the isometries.
- The Classical Groups of $\mathbb{C}^2$, for the groups that preserve each form.
- The Duo-Complex Bridge: The Three Canonical Forms and Their Tensor Powers, for the three forms carried to tensor powers, where $H$ gives the metric, $B$ the norm and $\omega$ the entanglement pairing.
- Bilinear Forms, for the general theory of symmetric and alternating forms.
- The Sesquilinear Structure of the Endomorphism Algebra, for the general theory of Hermitian forms.