Introduction to the Spin-One Anti-Hermitian Subspace

Introduction

This entry reads the last of the three carriers: the vector part, the anti-Hermitian space that closes it, and the object a boson carries at each event. The mathematics is one article and its two companions, and its answer is the simplest of the three in one respect and the richest in another. The carrier is the vector part, of real dimension three, and it is not a family: there is one triple, it depends on no axis, and every rotation of the frame is a symmetry of the whole structure. What makes the case rich is that the triple is not closed, and its closure is the half of the algebra that the rest of the corpus calls the material sector.

The first claim is that the carrier is the vector part, of real dimension three. The subspace is $\mathfrak{k}=\mathbb{R}\vec e_x\oplus\mathbb{R}\vec e_y\oplus\mathbb{R}\vec e_z$, the real span of the three pure units. On it the adjoint action acts, the weights are $\pm1$ and $0$, the spin is $1$, and the identity $2s+1=\dim_\mathbb{R}$ reads $3=3$. A full turn acts as $+\mathrm{id}$ and the kernel of the map to the rotations, which is $\{\pm e_t\}$, acts trivially.

The second claim is that the action is two-sided and that this is what makes the weights integers. The generator of a rotation is $\tfrac12\vec u$ and it acts by the commutator, and the commutator consumes the axis twice, once on each side. The identity $[\tfrac12\vec u,\vec v]=\vec u\times\vec v$ states the cancellation in one line, and it is what turns the half of the spinor case into the integer of this one. The commutator with a unit axis annihilates the axis and quarter-turns the plane perpendicular to it.

The third claim is that the triple is not carried onto itself by the whole group. A rotation preserves it and a boost does not: a boost along the axis sends $\vec u$ to $\cosh\phi\,\vec u-i\sinh\phi$, so the image acquires a component along the negative-definite line $\mathbb{R}(ie_t)$. This is the same failure of invariance that the spin-half carrier showed, and it is repaired the same way, by a closure.

The fourth claim is that the closure is the material sector. The subspace $\mathbb{B}_\Phi=\mathbb{R}(ie_t)\oplus\mathfrak{k}$ is of real dimension four, it is carried onto itself by the whole group, and the direction that was added is the central generator $ie_t$ and not the identity. The identity $e_t$ lies in the other half. So the closure is not the triple with a label attached: it is the triple with one more generator, and the group needs that generator to close.

The fifth claim is that the two invariants are real on the closure, and that this is what the sector buys. The norm is $N(i\lambda+\vec v)=\lvert\vec v\rvert^{2}-\lambda^{2}$, a real form with the signs $(3,1)$, positive definite on the carrier and negative definite on the added line. The square of the vector part is $B^{2}=\lvert\vec v\rvert^{2}$, non-negative, so the spectrally degenerate locus of the sector is the added line alone. One invariant measures the mass and the other measures the motion, both are real numbers, and no other carrier has that.

The sixth claim is that the carrier is the only one that lives entirely in the doubly regular region. Every nonzero element of the triple has $N\neq0$ and $B^{2}\neq0$, so the triple without its origin is contained in $\Sigma_{11}$, the region the corpus names mass in motion. The centre is confined to $\Sigma_{10}$ and the plane to $\Sigma_{01}$; the triple is the third and the last of the three, and it completes the sequence in which the two predicates are repaired one at a time.

The first objection is that a three-dimensional carrier cannot carry a three-state object and an algebra at once. It can, and the reason is the whole point of the case: the carrier is closed under the commutator, so it is a Lie algebra, and it is the Lie algebra of the group that acts on it. The representation is then the adjoint, and the three states are the algebra's own weight basis. This is the only one of the three carriers for which the carrier and the algebra of the action are the same object.

The second objection is that a Lie algebra is not a space a field takes values in. It is exactly such a space, and the corpus uses it that way: the gauge field of a non-abelian theory takes values in the Lie algebra of its group, because the connection is a one-form with values in that algebra and the field strength is its curvature. The objection describes the standard construction and not a defect of the carrier.

The third objection is that the corpus calls the material sector the third carrier, of real dimension four, while this entry calls the triple the carrier, of real dimension three. Both namings occur and both are correct at their own level. The carrier of the spin is the triple: three dimensions, three weights, the identity $2s+1=3$. The carrier of the amplitude is the sector: a field's amplitude is a four-vector, and a four-vector needs its time slot. The difference between the two is one real dimension, and it is exactly the dimension that the field equation's constraint removes. The entry states both and keeps them apart in a section of their own.

The letters. The algebra is $\mathbb{B}$, real of dimension eight, units $e_t,\vec e_x,\vec e_y,\vec e_z$ with $e_t$ the identity and $\vec e_k^{2}=-e_t$; the central imaginary is $i$; an element is $\tilde Q=Q_te_t+\vec Q$ with complex coefficients $Q_\mu=q_\mu+iq'_\mu$. The pure part is $\mathfrak{k}$, an element of it is written $\vec v$ with $\lvert\vec v\rvert^{2}=v_x^{2}+v_y^{2}+v_z^{2}$, and the cross product is $\times$. The star ${}^{*}$ is the Hermitian conjugation, the bar $\bar{\cdot}$ the coefficient conjugation, and ${}^{\natural}$ the natural conjugation. The norm is $N(\tilde Q)=Q_t^{2}+Q_x^{2}+Q_y^{2}+Q_z^{2}$, $B$ is its polar form, and $B^{2}=Q_x^{2}+Q_y^{2}+Q_z^{2}$ is the square of the vector part. The sectors are $\mathbb{B}_\Phi=\{X^{*}=-X\}$ and $\mathbb{B}_\Psi=\{X^{*}=X\}$. A rotation generator is $A(\vec u)=\tfrac12\vec u$; a rotor is $\tilde u_{\mathrm r}=\cos\tfrac{\theta}{2}e_t+\sin\tfrac{\theta}{2}\vec u$; a Lorentz rotor is $\tilde U_{\mathrm{lor}}$ with $N(\tilde U_{\mathrm{lor}})=1$, and a boost is $\cosh\tfrac{\phi}{2}e_t+i\sinh\tfrac{\phi}{2}\vec u_{\mathrm b}$. The four sectors of the two predicates are written $\Sigma_{ij}$.

Boundaries. The subspace, the closure and the weights are Introduction to the Spin-One Anti-Hermitian Subspace; the norm, the signs and the cone are The Spin-One Anti-Hermitian Subspace and N Norm Regularity; the predicate, the spectrum and the degenerate line are The Spin-One Anti-Hermitian Subspace and Spectral Regularity; the four sectors inside the closure are The Spin-One Anti-Hermitian Subspace and the Sectors; the list of three carriers and the closure rules are From the Spin to the Carriers and the Closures; the bracket and its Lie algebra are Lie Algebras of Sesqualgebras; the product's two halves are The Vector Parts of the Products over the Biquaternion Algebra; the rotor and the double cover are Biquaternion Rotations and Lorentz Transformations. The physics is taken from Four-Vectors and the Interval, Integer-Spin Quantization and the Adjoint Action on the Material Sector in Biquaternionic Form, The Proca Equation: Massive Spin 1 in Biquaternionic Form, The Third Polarization and the Self-Dual Split of the Massive Field Strength, The Self-Dual and Anti-Self-Dual Split: Spin 1 from the Bivector Plane, The Generalized Bloch Ball for Spin 1, Maxwell's Equations in the Biquaternionic Formulation, Mass in Motion: The Region of Both Invariants, The Gauge Principle in Biquaternionic Form and Why the Biquaternion Algebra Reaches SU(2) but Not SU(3). No dynamics, no gauge principle and no quantization appear below.

The Source Table

mathematical content source article
the vector part, the adjoint action, the closure, the generator and the three weights Introduction to the Spin-One Anti-Hermitian Subspace
the norm $\lvert\vec v\rvert^{2}-\lambda^{2}$, the signs $(3,1)$, the orthogonality and the null cone The Spin-One Anti-Hermitian Subspace and N Norm Regularity
the square $B^{2}=\lvert\vec v\rvert^{2}$, the spectrum $i(\lambda\pm\lvert\vec v\rvert)$ and the degenerate line The Spin-One Anti-Hermitian Subspace and Spectral Regularity
the four sectors inside the closure and the reach of the carrier The Spin-One Anti-Hermitian Subspace and the Sectors of N and Spectral Regularity
the list of three carriers, the criterion of the weight string and the four closure rules From the Spin to the Carriers and the Closures
the bracket, the Lie algebra $\mathfrak{u}(2)=\mathfrak{u}(1)\oplus\mathfrak{su}(2)$ and the abelian ideal Lie Algebras of Sesqualgebras
the commutator and the anticommutator of two pure products, and the split of the product The Vector Parts of the Products over the Biquaternion Algebra
the rotor, the two-to-one cover and the action on the pure part Biquaternion Rotations and Lorentz Transformations

The four rows that own the carrier. The first three state the subspace, the two predicates and the weights. The rest supply the structure that the physics uses: the place of the carrier in the list of three, the Lie algebra that makes it the algebra of the action, the split of a product into two halves, and the rotor.

The Results in Brief

The carrier. The vector part $\mathfrak{k}=\mathbb{R}\vec e_x\oplus\mathbb{R}\vec e_y\oplus\mathbb{R}\vec e_z$, of real dimension $3$, the space of real pure elements.

The action. $\operatorname{Ad}_U(X)=UXU^{*}$ at a Lorentz rotor with $N(U)=1$. It preserves $\mathfrak{k}$ when $U$ is a rotation and does not preserve it otherwise.

The generator. $A(\vec u)=\tfrac12\vec u$, acting by the commutator, with $[\tfrac12\vec u,\vec v]=\vec u\times\vec v$.

The weights. $\{+1,0,-1\}$ on the complexified carrier, the eigenvalue $0$ on the axis and the pair on the perpendicular plane.

The full turn. $\exp(2\pi C(\vec u))=\mathrm{id}$ and $-\mathrm{id}$ acts as the identity, so the kernel $\{\pm e_t\}$ is invisible and $(-1)^{2s}=+1$.

The closure. The material sector $\mathbb{B}_\Phi=\mathbb{R}(ie_t)\oplus\mathfrak{k}$, of real dimension $4$, invariant under the whole group and a Lie algebra under the bracket.

The two invariants. $N(i\lambda+\vec v)=\lvert\vec v\rvert^{2}-\lambda^{2}$ and $B^{2}(i\lambda+\vec v)=\lvert\vec v\rvert^{2}$; the first has the signs $(3,1)$ and the second is non-negative.

The spectrum. $\operatorname{spec}(i\lambda+\vec v)=\{i(\lambda+\lvert\vec v\rvert),\,i(\lambda-\lvert\vec v\rvert)\}$, purely imaginary, degenerate exactly on the line $\mathbb{R}(ie_t)$.

The reach. $\mathfrak{k}\setminus\{0\}\subseteq\Sigma_{11}$; the closure meets all four sectors.

The spin content. $\mathfrak{k}$ is one irreducible representation of weight string $(1,0,-1)$; the closure is $1\oplus0$ under the rotation subgroup, the extra $0$ being the added line.

The Triple

The subspace. The carrier is the real span of the three pure units,

$$ \mathfrak{k}=\mathbb{R}\vec e_x\oplus\mathbb{R}\vec e_y\oplus\mathbb{R}\vec e_z =\{\,\vec v:\vec v^{*}=-\vec v,\ \vec v\ \text{real}\,\}, \qquad \dim_\mathbb{R}\mathfrak{k}=3 . $$

The two descriptions agree: a pure element is anti-Hermitian exactly when its coefficients are real, because the star reverses the coefficient and the vector part. So the triple is the real pure part, and it is the intersection of the pure part with the material sector.

What it is closed under. The three units anticommute in pairs and each squares to $-e_t$, so the span is closed under the commutator and not under the product. The product rule is the whole of the structure,

$$ \vec a\,\vec b=-\,\vec a\cdot\vec b+\vec a\times\vec b , $$

and the two terms have different characters: the dot product is symmetric and lands on the scalar line, the cross product is antisymmetric and lands back in the triple. So the square of an element of the triple is a real number, $\vec v^{2}=-\lvert\vec v\rvert^{2}e_t$, and the product of two elements of the triple is a pure element that is not real unless the two are parallel.

The three dimensions. Nothing in the definition singles out a direction. The triple has no preferred axis, no parameter and no family: the group acts on it transitively away from the origin, and the orbits are the spheres $\lvert\vec v\rvert=r$. This is the structural difference from the spin-half carrier, whose every statement carried the axis $\vec u$ as a parameter, and it is the reason the three articles of this case are shorter than the two of that one.

The pairing with the centre. The centre of the algebra is the fixed space of the adjoint action, and the centre meets the triple at the origin only. So the triple carries no central direction and the centre carries no vector part: the two carriers are disjoint, and the closure below is what puts a central direction next to the triple without confusing the two.

The polar form on the triple. On the triple the polar form of the norm is the Euclidean inner product,

$$ B(\vec v,\vec w)=\vec v\cdot\vec w , $$

because the norm there is $\lvert\vec v\rvert^{2}$ and the polar form is half the difference of the squares of the sum and of the pieces. So the triple is a Euclidean three-space with the group acting by rotations, and the invariant of that action is the length.

The Action and the Weight Basis

The group and its action. The elements of unit quaternionic norm form a group, and the action on the algebra is

$$ X\longmapsto \operatorname{Ad}_U(X)=U\,X\,U^{*}, \qquad N(U)=1 . $$

**On the triple the action is a rotation when $U$ is a real unit quaternion and something else when $U$ carries the central imaginary. The real unit quaternions are the rotation rotors, and the map from them to the rotations is onto with kernel $\{\pm e_t\}$.

The generator. A rotation rotor is $\tilde u_{\mathrm r}=\exp\bigl(\theta A(\vec u)\bigr)$ with $A(\vec u)=\tfrac12\vec u$, and the derived operator on the triple is the commutator,

$$ \rho_{\mathrm{rep}}\bigl(A(\vec u)\bigr)(\vec v)=\bigl[\tfrac12\vec u,\vec v\bigr]. $$

The identity that removes the computation. The commutator of a half axis with a vector is the cross product with the whole axis,

$$ \bigl[\tfrac12\vec u,\vec v\bigr]=\tfrac12\bigl(\vec u\vec v-\vec v\vec u\bigr)=\vec u\times\vec v , $$

the residual of which is $0$ over one hundred random pairs. The factor one half in the generator is cancelled by the factor two in the commutator. This is the entire computation of the case, and it holds at every axis.

The eigenvectors. The cross product with a unit vector annihilates the axis and quarter-turns the plane perpendicular to it. With $\vec u_2,\vec u_3$ an oriented orthonormal pair completing the axis, the three complex combinations

$$ \varepsilon_+=\frac{\vec u_2+i\vec u_3}{\sqrt2}, \qquad \varepsilon_-=\frac{\vec u_2-i\vec u_3}{\sqrt2}, \qquad \varepsilon_0=\vec u $$

satisfy

$$ \bigl[\tfrac12\vec u,\varepsilon_+\bigr]=-i\,\varepsilon_+, \qquad \bigl[\tfrac12\vec u,\varepsilon_-\bigr]=+i\,\varepsilon_-, \qquad \bigl[\tfrac12\vec u,\varepsilon_0\bigr]=0 . $$

The three vectors are the circular and longitudinal directions of the frame whose third axis is $\vec u$: $\varepsilon_\pm$ are the two circular directions of the transverse plane and $\varepsilon_0$ is the axis itself.

The weight operator. The generator is a real element, so the commutator is a real antisymmetric operator and its eigenvalues are imaginary. The observable is the Hermitian operator obtained by one multiplication by the central imaginary,

$$ \tilde S_3=i\bigl[\tfrac12\vec u,\ \cdot\ \bigr], $$

The weight operator is Hermitian. It is Hermitian on the complexified carrier, verified to $6.3\times10^{-16}$, and whose eigenvalues are the weights,

$$ \tilde S_3\varepsilon_+=\varepsilon_+, \qquad \tilde S_3\varepsilon_-=-\varepsilon_-, \qquad \tilde S_3\varepsilon_0=0 . $$

So the weight string is $(+1,-1,0)$, the largest weight is $1$, and the identity reads

$$ 2s+1=3=\dim_\mathbb{R}\mathfrak{k}. $$

Why the weight operator carries the central imaginary. The generator $A(\vec u)$ is a material element, and a material element generates a real flow. A measurement reads a Hermitian operator, and the operator has to be multiplied by $i$ to become one. So the weight operator, the object a spectroscopist diagonalises, lies in the informational sector even when the generator lies in the material one. This is the corpus's standing convention that an observable is the imaginary multiple of a material element, read on the smallest example; the spin-zero and spin-half entries met the same multiplication by $i$ in the double cover.

Which circular vector carries which weight is a convention. The identity above fixes $\varepsilon_+$ as the vector carrying $+1$, which is a choice of handedness of the pair $(\vec u_2,\vec u_3)$. The corpus does not agree with itself on the sign throughout: Integer Spin Quantization and the Adjoint Action writes $\tilde S_{\mathrm{adj},3}\varepsilon_+ = +\varepsilon_+$ with the same $\varepsilon_+$, and it reaches the same spectrum by the opposite sign in the operator. Nothing depends on the choice except the names left and right.

The full turn. Four quarter turns are a full turn, so the operator of a full turn is the identity,

$$ \exp\bigl(2\pi\,C(\vec u)\bigr)=\mathrm{id}, \qquad C(\vec u)(\vec v)=\vec u\times\vec v , $$

and the same holds at the level of the group: the rotor of a full turn is $-e_t$, and $-e_t$ acts on the triple as the identity, to $8.9\times10^{-16}$ over one hundred vectors. So the two signs of the rotor are indistinguishable here and $(-1)^{2s}=(-1)^{2}=+1$ agrees with the computation.

The half turn is not the identity. The case is not that every angle acts trivially: at $\pi$ the transverse vector $\vec u_2$ is sent to $-\vec u_2$, at Euclidean distance $2$. What is trivial is the whole turns and the sign of the rotor, and the two statements are different.

Why the Action Is Two-Sided, and What It Costs

The two actions. The same generator acts on the spin-half carrier by a single multiplication and on the triple by the commutator. The two spectra are $\{\pm\tfrac12\}$ and $\{\pm1,0\}$, and the ratio is two.

The reason, in one line. The commutator consumes the axis twice, once on each side of the product, and a single multiplication consumes it once. In the identity $[\tfrac12\vec u,\vec v]=\vec u\times\vec v$ the two consumptions appear as the cancellation of the one half: half of the generator on each side equals the whole generator on one side.

What it costs. A two-sided action requires the acting element and its conjugate on either side, so it needs the inverse, so it needs the norm to be invertible. A one-sided action needs only a product. The triple is therefore the carrier on which the group acts by an automorphism, and the plane is the carrier on which it acts by a multiplication from one side only.

Where the automorphism stops. The action $\operatorname{Ad}_U$ is multiplicative only when $U^{*}U=e_t$, and that holds for the rotation rotors and fails for the boosts. For a boost of rapidity $\phi$ along $\vec u_{\mathrm b}$,

$$ U^{*}U=\cosh\phi\,e_t+i\sinh\phi\,\vec u_{\mathrm b}, $$

verified to $6.3\times10^{-16}$, so $\operatorname{Ad}_B(XY)$ and $\operatorname{Ad}_B(X)\operatorname{Ad}_B(Y)$ differ by that factor and the map is not an algebra automorphism. The action is nevertheless a group action and it preserves the sector and the norm. The distinction is worth stating because the word conjugation suggests an automorphism and, for a boost, the suggestion is false.

Why the Triple Is Not a Carrier for the Group

The test. A subspace is a carrier when the action of the group carries it onto itself. On the triple the action is a rotation for a rotation rotor and something else for a boost, and the case is decided in one computation.

A rotation keeps the triple. A rotation rotor is real and of unit norm, so its star is its inverse and the action is the ordinary conjugation. The conjugate of a real pure element is a real pure element, the scalar part stays zero, and the residual over one hundred random vectors is $2.2\times10^{-16}$. So the rotation subgroup acts on the triple and its action is the rotation of the three-space.

A boost does not. A boost rotor is Hermitian of unit norm, $U_{\mathrm b}=\cosh\tfrac{\phi}{2}e_t+i\sinh\tfrac{\phi}{2}\vec u_{\mathrm b}$, so $U_{\mathrm b}^{*}=U_{\mathrm b}$ and the action is $X\mapsto U_{\mathrm b}XU_{\mathrm b}$, a two-sided multiplication by the same element. On the transverse directions nothing happens,

$$ \operatorname{Ad}_{U_{\mathrm b}}(\vec e_x)=\vec e_x , \qquad \operatorname{Ad}_{U_{\mathrm b}}(\vec e_y)=\vec e_y , $$

when the boost is along $\vec e_z$, verified to $6.7\times10^{-16}$. On the axis the element leaves the triple:

$$ \operatorname{Ad}_{U_{\mathrm b}}(\vec e_z)=\cosh\phi\,\vec e_z-i\sinh\phi , \qquad \operatorname{Ad}_{U_{\mathrm b}}(ie_t)=i\cosh\phi\,e_t-\sinh\phi\,\vec e_z , $$

both verified to $4.4\times10^{-16}$ over one hundred rapidities. The first line is the failure: a pure vector is sent to a sum of a pure vector and a multiple of $ie_t$, so the image is no longer in the triple but it is still in the sector. The two lines together are the boost of the four-vector basis, with the time slot and the axis slot mixing by the hyperbolic rotation.

The size of the failure. A vector parallel to the boost axis of length $r$ acquires a scalar of magnitude $r\sinh\phi$ and keeps $r\cosh\phi$ along the axis. The failure is therefore not a small effect and it is not removable by a choice of frame: it is proportional to the rapidity, and at a boost the triple is broken by an amount that grows without bound.

The repair. The image of the action is contained in the span of the triple and the line $\mathbb{R}(ie_t)$, and that span is closed. So the closure of the carrier is the four-dimensional space, and it is the smallest subspace containing the triple and stable under the whole group. This is the same mechanism as in the spin-half case, where the image of a transverse multiplication by an axis forced the addition of two real dimensions; here the addition is one real dimension and the resulting space is the material sector.

The Closure: the Material Sector

The space. The closure is

$$ \mathbb{B}_\Phi=\{\,X:X^{*}=-X\,\}=\mathbb{R}(ie_t)\oplus\mathfrak{k}, \qquad \dim_\mathbb{R}\mathbb{B}_\Phi=1+3=4 , $$

the anti-Hermitian subspace. It is invariant under the whole group: for $X^{*}=-X$ and any rotor,

$$ (UXU^{*})^{*}=U\,X^{*}\,U^{*}=-\,UXU^{*}, $$

so the conjugate is anti-Hermitian again, verified to $8.9\times10^{-16}$ over one hundred random sector elements and boosts. The invariance is a one-line consequence of the star being an anti-automorphism, and it holds for every rotor and not only for the rotations.

The added direction is a generator and not the identity. The line that completes the triple is $\mathbb{R}(ie_t)$, and $e_t$ is not in it. This matters for two reasons. The first is sectorial: $e_t$ is Hermitian and lies in $\mathbb{B}_\Psi$, while $ie_t$ is anti-Hermitian and lies in the closure, so the two lines are in different halves and a reader who completes the triple with the identity has left the sector. The second is structural: $ie_t$ is the central element that generates the abelian factor, the element that squares to minus the identity and that a rotation cannot move, while $e_t$ is the identity that acts trivially on everything. A closure built with the identity would add nothing; the closure built with $ie_t$ adds the time direction.

The bracket closes. For two anti-Hermitian elements the commutator is anti-Hermitian and the anticommutator is Hermitian, exactly:

$$ [\mathbb{B}_\Phi,\mathbb{B}_\Phi]\subseteq\mathbb{B}_\Phi, \qquad \{\mathbb{B}_\Phi,\mathbb{B}_\Phi\}\subseteq\mathbb{B}_\Psi , $$

residual $0$ over one hundred random pairs for each. So the closure is a Lie algebra under the bracket and the other half is not, and a space that is read with the bracket has to be this one.

The Lie algebra it is. With the bracket the closure is $\mathfrak{u}(2)$, and the split of the bracket is

$$ \mathbb{B}_\Phi=\mathfrak{u}(1)\oplus\mathfrak{su}(2), \qquad \mathfrak{u}(1)=\mathbb{R}(ie_t), \qquad \mathfrak{su}(2)=\mathfrak{k}, $$

of dimensions $1+3=4$. The line is an abelian ideal, fixed pointwise by the adjoint action of the rotations, and the triple is the non-abelian summand on which the bracket does not vanish. The bracket of two sector elements in coordinates is twice the cross product of their space parts, with the time slot contributing nothing.

The spin content of the closure. Under the rotation subgroup the two summands are irreducible and of different weights: the triple carries the weight string $(1,0,-1)$ and the line carries the weight $0$ alone, because a rotation fixes it. So

$$ \mathbb{B}_\Phi\;\cong\;\mathbf{1}\oplus\mathbf{0} $$

as a representation of the rotation group, the bold numeral being the spin. The closure adds a singlet and not a second triplet, and the singlet is the time direction.

Why the sector and not the other half. The other half $\mathbb{B}_\Psi=\mathbb{R}e_t\oplus i\mathfrak{k}$ has the same shape and the same dimension, and the triple does not lie in it: $\mathfrak{k}\cap\mathbb{B}_\Psi=\{0\}$. So the half that contains the carrier is forced, and there is no choice to make. The triple $i\mathfrak{k}$ that lies in $\mathbb{B}_\Psi$ is imaginary, and the carrier is real.

The Two Namings: Three Dimensions and Four

The two objects. The carrier of the spin is the triple, of real dimension three, and the identity $2s+1=\dim_\mathbb{R}$ is the test that fixes it. The carrier of the amplitude is the sector, of real dimension four, and the test that fixes it is that a field's amplitude is a four-vector. The two objects are nested, $\mathfrak{k}\subset\mathbb{B}_\Phi$, and the difference is one real dimension.

Why the naming is a convention with a warrant. The mathematics names $\mathbb{B}_\Phi$ for the carrier and states that no number depends on the choice: with the other half named, the types, the weights $\{0,\pm1\}$ and the spin would be unchanged. The warrant is that $\mathbb{B}_\Phi$ contains the carrier and is a Lie algebra under the bracket, and a momentum is a four-vector that lies in the half whatever the spin of its field, while an amplitude lies in the carrier that the spin fixes. For the vector field the two coincide because a vector amplitude is a four-vector, and the naming follows the amplitude.

Why this entry names the triple. The three-state count is a statement about the spin, and it is exact: three weights, three dimensions, three polarizations. A four-dimensional space is the carrier of the spin and of a scalar, and the scalar is the time slot. So the entry keeps the two namings and states the difference, rather than choosing one and hiding the other.

The difference is exactly the constraint. The Proca equation implies the Lorenz condition, and the condition removes one real degree of freedom, which is the time slot: the singlet that the adjoint action annihilates. What remains is the triple. So the four-dimensional carrier is the triple together with the degree of freedom that the equation then eliminates, and the arithmetic is

$$ \dim_\mathbb{R}\mathbb{B}_\Phi-1=3=\dim_\mathbb{R}\mathfrak{k}. $$

And then the mass removes another. For the massless field the gauge freedom removes one more, and two polarizations remain. So the sequence of counts is $4\to3\to2$: four components in the potential, three physical states for a massive boson, two for a massless one. Each step is an equation rather than a convention, and the entry states the chain because it is the cleanest statement of what the carrier and the closure are for.

The sector is not the carrier's double. A reader might think that the fourth dimension is a second copy of the carrier. It is not: the added dimension is a different object, of different signs, carrying a different weight string, and fixed by the rotations. The closure adds a singlet and the carrier is the triplet; the two are not interchangeable and the entry names them separately everywhere.

The reach is different for the two. The distinction is not only dimensional. The triple without its origin lies entirely in $\Sigma_{11}$, while the sector meets all four sectors: the line away from the origin is in $\Sigma_{10}$, the cone away from the origin is in $\Sigma_{01}$, the origin is in $\Sigma_{00}$, and the interior is in $\Sigma_{11}$. So the triple is the carrier of one regime and the sector is the carrier of the classification. That is the subject of a later section.

The Two Numbers: the Norm and the Spectrum

The norm on the closure. For real $\lambda$ and real $\vec v$ the norm is

$$ N(i\lambda+\vec v)=Q_t^{2}+Q_x^{2}+Q_y^{2}+Q_z^{2}=(i\lambda)^{2}+\lvert\vec v\rvert^{2}=\lvert\vec v\rvert^{2}-\lambda^{2}, $$

a real number, residual $0$ over one hundred random sector elements. On the triple it is $\lvert\vec v\rvert^{2}$, positive definite, and its least value on the unit sphere is $1$ exactly. On the added line it is $-\lambda^{2}$, negative definite, and on the imaginary triple $i\mathfrak{k}$ it is $-\lvert\vec v\rvert^{2}$, also negative definite, to $0$ over one hundred draws.

The two summands are orthogonal. The polar form couples them to nothing,

$$ B(i\lambda,\vec v)=0, \qquad N(i\lambda+\vec v)=N(i\lambda)+N(\vec v), $$

so the norm of a sector element is the sum of the norms of its two pieces and the two pieces never interfere.

The signs, and where the indefiniteness comes from. The form has the signs $(3,1)$: positive on the three dimensions of the triple and negative on the one dimension of the line. Neither summand is indefinite. Each is definite, one with each sign, and the indefiniteness of the closure is created by adding them. This is worth stating plainly, because the Lorentzian sign pattern is usually presented as a property of the whole space and here it is visibly a property of an addition.

The square of the vector part. The other invariant is

$$ B^{2}(i\lambda+\vec v)=Q_x^{2}+Q_y^{2}+Q_z^{2}=\lvert\vec v\rvert^{2}, $$

a non-negative real number, residual $0$ over one hundred random elements. It does not see the time slot at all. So the spectrally degenerate locus of the closure is the set where the vector part vanishes, which is the added line:

$$ \mathbb{B}_\Phi\cap\{B^{2}=0\}=\mathbb{R}(ie_t), $$

of real dimension one.

The null cone. The norm vanishes where the two lengths are equal,

$$ \mathcal N\cap\mathbb{B}_\Phi=\{i\lambda+\vec v:\lvert\vec v\rvert^{2}=\lambda^{2}\}, $$

a cone of real dimension three with two nappes, each nappe meeting each sphere of the closure in a sphere of dimension two. The cone meets the triple at the origin and the line at the origin, and it separates the region of positive norm from the region of negative norm.

The spectrum. The two eigenvalues of a sector element are

$$ \operatorname{spec}(i\lambda+\vec v)=\bigl\{i(\lambda+\lvert\vec v\rvert),\,i(\lambda-\lvert\vec v\rvert)\bigr\}, $$

verified to $1.8\times10^{-15}$ over one hundred random elements. Both lie on the imaginary axis and are symmetric about $i\lambda$. On the carrier, with $\lambda=0$, the spectrum is $\{\pm i\lvert\vec v\rvert\}$ and the two eigenvalues are distinct for every nonzero vector, so the carrier is spectrally regular away from the origin.

Why the spectrum is imaginary. The closure is defined by $X^{*}=-X$, and the two eigenvalues are $Q_t\pm iB$ with $Q_t=i\lambda$ and $B=\lvert\vec v\rvert$. The scalar part supplies the $i\lambda$ and the vector part supplies a real $B$, so both corrections are imaginary. The reality of $\vec v$ is what makes $B$ real; on a general complex element $B$ need not be real and the eigenvalues need not be imaginary.

The two loci do not meet except at the origin. A sector element is on the cone when $\lvert\vec v\rvert=\lvert\lambda\rvert$ and on the degenerate line when $\vec v=0$; both hold only at the origin. So the cone and the degenerate line are a cone and a line crossing in one point, and neither contains the other. The plane of the spin-half case had the opposite relation: the whole plane was on the cone and the degenerate locus was one point of it.

The two numbers side by side. The residuals of the identities of this section, on one hundred elements each:

identity residual
$N(i\lambda+\vec v)=\lvert\vec v\rvert^{2}-\lambda^{2}$ $0$
$B^{2}(i\lambda+\vec v)=\lvert\vec v\rvert^{2}$ $0$
$B(i\lambda,\vec v)=0$ $0$
$\operatorname{spec}(i\lambda+\vec v)=i(\lambda\pm\lvert\vec v\rvert)$ $1.8\times10^{-15}$
$N(i\vec k)=-\lvert\vec v\rvert^{2}$ $0$
least $N$ on the unit sphere of $\mathfrak{k}$ $1$, exactly

The Signs, and the Sign of One Term

One sign and its cause. The interval of the closure is indefinite because one term carries $i$,

$$ N(i\lambda+\vec v)=-\lambda^{2}+\lvert\vec v\rvert^{2}, $$

and the $i$ is what makes the time term negative while the three space terms are positive. With a real $\lambda$ in place of $i\lambda$ the same space would carry a Euclidean four-coordinate geometry and no cone at all.

The convention, stated as a convention. The placement of the central imaginary on the time slot is a convention of the corpus and not a theorem of the algebra. The other half of the algebra makes the opposite placement, $\mathbb{B}_\Psi=\mathbb{R}e_t\oplus i\mathfrak{k}$ with elements $\lambda+i\vec v$, and its norm is

$$ N(\lambda+i\vec v)=\lambda^{2}-\lvert\vec v\rvert^{2}, $$

of the signs $(1,3)$, with a real spectrum $\lambda\pm\lvert\vec v\rvert$ instead of an imaginary one. Both halves are equally indefinite and both carry a cone; what differs is which triple is real, and the carrier is the real one. The corpus chooses the anti-Hermitian half because the triple is real there, and it records the choice.

What the convention does not change. The dimensions, the bracket, the weights, the invariance of the sector under the group and the two-to-one cover are the same on both halves. What the convention changes is the sign of the norm, and with it the causal reading. So the entry states the interval as a consequence of the placement rather than as a consequence of the algebra.

The physics of the choice. The Lorentzian signs are what make the light cone exist, and the light cone is what makes the causal structure, the mass shell and the speed limit exist. All of that follows from one sign, and the sign follows from one placement of $i$. This is the closest the framework comes to deriving the signs, and the entry states the limit: it derives the signs from the convention, not the convention from anything.

The spectrum on the two halves. The predicate is the same on both halves, since it is decided by $B^{2}$ and that is $\lvert\vec v\rvert^{2}$ on the one and $-\lvert\vec v\rvert^{2}$ on the other, vanishing in both cases exactly where the vector part does. What moves is the location of the spectrum: imaginary on the anti-Hermitian half, real on the Hermitian one. So the predicate that classifies the element does not read the axis on which its eigenvalues lie, and the two halves are the sharpest illustration of that separation in the corpus.

The Three Carriers and the Three Sectors

The three carriers. The two predicates cut the algebra into four regions, and the three carriers occupy three of them, one each, away from the origin. The centre is the fixed space of the adjoint action, of real dimension two; the plane is one-sided, of real dimension two; the triple is two-sided, of real dimension three. The table repeats the facts proved in the three companion entries and places them side by side.

carrier subspace $\dim_\mathbb{R}$ action weights spin region away from the origin
the centre $\mathbb{C}e_t$ $2$ adjoint, trivial $0$ $0$ $\Sigma_{10}$
the plane $\mathbb{C}\tilde\Pi_+(\vec u)$ $2$ left multiplication $\pm\tfrac12$ $\tfrac12$ $\Sigma_{01}$
the triple $\mathfrak{k}$ $3$ adjoint $\pm1,0$ $1$ $\Sigma_{11}$

Why the assignment is forced. The centre has $B^{2}=0$ identically, so no point of it can be in a region with $B^{2}\neq0$. The plane has $N=0$ identically, so no point of it can be in a region with $N\neq0$. The triple has both quantities positive at every nonzero point, so it is contained in $\Sigma_{11}$ and in no other. The three assignments follow from the two identities $B^{2}=0$ on the centre and $N=0$ on the plane, and from no further computation.

The sequence. The order is the order of repair. The centre has the spectrum and loses the norm. The plane has the norm and loses the spectrum. The triple has both. So the list of three carriers is also a list of three states of the pair of predicates, and the third entry closes the list because no further repair is available: a carrier cannot have more than both.

Where the sector sits in the table. The sector is not a fourth carrier. It is the closure of the third, and it meets all four regions: the line away from the origin in $\Sigma_{10}$, the cone in $\Sigma_{01}$, the origin in $\Sigma_{00}$, the interior in $\Sigma_{11}$. So the sector is the object that spans the whole classification, while each of the three carriers is confined to one region. This is the precise sense in which the sector is the carrier of the classification and the triple is the carrier of the spin.

The origin. All three carriers contain the origin, and the origin is in $\Sigma_{00}$. So the statement "the centre is in $\Sigma_{10}$" is meant away from the origin, and the entry states it that way throughout. Nothing in the case depends on the value at the origin, because the origin is the zero field and carries no state.

What the table does not say. It does not say that the three carriers exhaust the subspaces of the algebra. It says that they exhaust the ones that are invariant and irreducible under one of the two actions, which is the definition. The ideals, for instance, are two more subspaces of real dimension four each, and they are carriers of the left action but reducible for the adjoint; they belong to the spin-half entry and the four ideals of the algebra are treated with the bispinors.

The Interpretations

  1. Spin one is the adjoint representation, and this is why its carrier is special. For the other two carriers the group acts on something other than itself: on a line by the trivial action, on a plane by a multiplication from one side. For the triple the group acts on its own Lie algebra by conjugation, so the representation is the adjoint, and the carrier is the same object as the algebra of the action. The corpus's three carriers are therefore not three examples of one construction: the third is the construction in which the carrier and the symmetry coincide, and that coincidence is what "spin one" means in this framework.

  2. The three states are the triad of a frame. The weight basis of the carrier is $\varepsilon_+,\varepsilon_-,\varepsilon_0$: two circular directions in the plane perpendicular to the axis and one along the axis. These are the two transverse polarizations and the longitudinal one, the same three directions the corpus's Integer Spin Quantization article builds for a spin-one field, and the same three that The Third Polarization and the Self-Dual Split counts at a fixed wavevector. The count $2s+1=3$ is not an abstract dimension count here; it is the number of independent directions a vector can have relative to an axis.

  3. The integer weight is the price of two-sidedness, and the boson does not pay the sign. The commutator halves nothing: the half in the generator is cancelled by the two sides, and the weights come out as integers. A full turn is $+\mathrm{id}$ and the kernel $\{\pm e_t\}$ of the two-to-one cover acts trivially, which is the statement $(-1)^{2s}=+1$. So a spin-one object does not distinguish a rotation from the rotation followed by a full turn, and the double cover is invisible to it. The spin-half entry pays the sign because its action is one-sided; the spin-one carrier does not, and the difference is exactly the factor of two in the commutator.

  4. A boost turns spin into time, and this is why spin is not a four-vector. Under a boost along the axis the pure vector $\vec e_z$ becomes $\cosh\phi\,\vec e_z-i\sinh\phi$: the spin direction keeps a spatial part and gains a time part. So the three spin states of one frame are not the three spin states of another, and the object that is covariant is the four-component amplitude in the sector, not the triple. This is the framework's version of the standard statement that the spin of a particle is defined in its rest frame and that a moving particle's polarization is a four-vector. The identity $\operatorname{Ad}_{U_{\mathrm b}}(\vec e_x)=\vec e_x$ for a boost along $z$ is the other half of the same statement: the transverse directions are untouched, and only the axis mixes with time.

  5. The rest frame is the spectrally degenerate locus. The degenerate locus of the sector is the line $\mathbb{R}(ie_t)$, and a momentum at rest is $i\,mc\,e_t$, which is on that line. So the eigenvalues of the momentum coalesce exactly when the particle is at rest, and the spectral degeneracy of the material sector is the rest frame. Conversely the region $\Sigma_{11}$, where the vector part does not vanish, is motion: this is the precise content of the corpus's name "mass in motion" for that region, and it is a statement about the spectrum and not a metaphor.

  6. Mass is the norm and motion is the square of the vector part. The two invariants of the sector are independent: $N=\lvert\vec v\rvert^{2}-\lambda^{2}$ reads the interval and $B^{2}=\lvert\vec v\rvert^{2}$ reads the motion. For a momentum $p=i\,p'_te_t+\vec p$ the first is $-m^{2}c^{2}$ and the second is $\lvert\vec p\rvert^{2}$, so the pair $(N,B^{2})$ is the pair (mass, momentum) in invariant form. No other carrier has two independent real invariants, because on the centre and on the plane one of the two vanishes identically. The four-region classification is therefore a classification of momenta, and it is available only here.

  7. The light cone needs the added direction, so masslessness is not in the carrier. A nonzero element of the triple has $N=\lvert\vec v\rvert^{2}>0$, so no point of the carrier is null and the carrier never touches the cone away from the origin. A null momentum needs both slots, $p=i\,E e_t+E\hat k$, with the two lengths equal. So the carrier of the spin cannot hold a massless state on its own, and the closure is what makes one available. The physics is the standard one: the momentum of a photon has a time component, and its spatial part alone is spacelike.

  8. The carrier is the algebra a gauge field takes values in. The closure is $\mathfrak{u}(2)$ and the carrier is its $\mathfrak{su}(2)$ summand, and a non-abelian connection is by construction a one-form with values in the Lie algebra of its group, with the field strength as its curvature. So the corpus's carrier of spin one is the same object that gauge theory uses for the gauge bosons of a non-abelian group: the three states are the three components of the algebra, and the adjoint action is the transformation law. This is the one place in the corpus where the carrier of a spin and the values of a connection are literally the same space, and it is why the spin-one entry is the natural place to read the gauge principle. The framework's own ceiling, that the algebra reaches $\mathfrak{su}(2)$ and not $\mathfrak{su}(3)$, is the limitation of that reading and is stated in Why the Biquaternion Algebra Reaches SU(2) but Not SU(3).

  9. The field strength leaves the carrier, and that is why $\vec E$ and $\vec H$ are different objects. The product of two sector elements splits into a commutator, which is anti-Hermitian and stays in the sector, and an anticommutator, which is Hermitian and leaves it. In the field strength $\tilde F=i\sqrt\epsilon\,\vec E-\sqrt\mu\,\vec H$ the magnetic part is the real vector part and the electric part is the imaginary one, so the two halves of the field sit in the two halves of the complexified carrier, one material and one informational. The carrier of the potential is the sector; the carrier of the field is the complexified vector part; the two are not the same space and the split of the product is the reason.

  10. The sector has no complex structure, so the field is neutral. Multiplication by the central imaginary carries the sector onto the other half, $i\mathbb{B}_\Phi=\mathbb{B}_\Psi$, so the sector is a real form and its scalars are real. A complex structure on the carrier is what a charge would need, and the only one available leaves the sector. So the spin-one carrier as such carries a real, neutral field, and a charged field needs the larger object in which the two halves are both present. This is the framework's version of the statement that a real vector field has no charge and that a charged one has two real components per mode.

  11. The constraint removes the singlet, and the mass removes one more. The chain of counts is $4\to3\to2$: four components in the potential, three physical states for a massive boson, two for a massless one. The first step is the Lorenz condition, which removes the time slot, that is, exactly the dimension that the closure added; the second is the gauge freedom of the massless case. So the relation between the carrier and the closure is not a bookkeeping device: the extra dimension is the one that the field equation deletes, and the deletion is what makes the four-component object into the three-state one.

  12. The action is an automorphism only for rotations, and that is the little group. The action $\operatorname{Ad}_U$ preserves the bracket when $U^{*}U=e_t$, which holds for the rotation rotors and fails for the boosts. So the rotations act on the carrier by algebra automorphisms and a boost does not. The subalgebra that a boost along $\vec u_{\mathrm b}$ leaves alone is the one that commutes with the rotor: the rotation about $\vec u_{\mathrm b}$ itself, since $[\vec u_{\mathrm b},U_{\mathrm b}]=0$ exactly, and not the two rotations about the transverse axes. So a moving frame keeps one circle of the three, and the algebra that a moving observer sees is a $\mathfrak{u}(1)$ inside the $\mathfrak{su}(2)$ of the rest frame. This is the framework's form of the statement that the compact part of the little group shrinks from the rotation group to the rotations about the momentum, and it explains why a boost cannot be used to define a spin basis.

  13. The three carriers are now complete, and the list has a last member for a reason. The centre, the plane and the triple exhaust the invariant irreducible subspaces, and they are the three possible responses to the pair of predicates: lose the norm, lose the spectrum, lose neither. Spin zero, spin half and spin one are therefore not three values chosen from a range: they are the three ways a subspace can relate to the two invariants, and the algebra has no fourth. This is the argument the corpus's Determination of the Possible Spins article makes from the inside, and the three entries of this sequence make it from the outside by exhibiting the spaces.

The Limits

No dynamics. The entry writes the carrier, the closure and the two invariants, and it writes no field equation. The Proca equation, the Lorenz condition and the mass shell are cited as the objects that use the carrier, and the derivation is The Proca Equation: Massive Spin 1 in Biquaternionic Form. Nothing here is derived from a Lagrangian and no solution is written.

The placement of the central imaginary is a convention. The signs $(3,1)$ are a consequence of writing the time slot as $i\lambda$ and not a consequence of the algebra. The other half of the algebra gives $(1,3)$ with the same construction, and the entry states the choice rather than deriving it. A reader who wants a Euclidean four-space can have one by moving the $i$, and the carrier of the spin is unaffected.

The identification of the triple with the three spin states is an interpretation. The algebra supplies a three-dimensional carrier with the weight string $(1,0,-1)$ and a group acting by the adjoint representation. That the three weights are the three polarizations of a vector particle is a physical reading, supported by the corpus's Proca and Maxwell articles, and it is not a theorem of the algebra.

The time slot is named by convention too. Nothing in the algebra distinguishes the scalar line from the three vector lines except the sign of the square; a reader could call the scalar line a spatial direction and the three vector lines the time and two spaces, and every identity would survive. So "the fourth direction is time" is the corpus's reading of the one negative direction, and it is a reading.

The carrier is not the field strength. The potential takes values in the sector and the field strength takes values in the complexified vector part, which is a larger space of real dimension six. The entry states the relation and does not derive the field strength from the carrier; the derivation is in the Proca and Maxwell entries.

No gauge principle. The entry notes that a non-abelian connection takes values in a Lie algebra and that the carrier is one, and it does not derive the gauge principle, the covariant derivative or the field strength from that observation. The corpus states its own ceiling on that programme in Why the Biquaternion Algebra Reaches SU(2) but Not SU(3), and the entry does not raise it.

No quantization. The weight operator is a classical Hermitian operator on a complexified real vector space, and its eigenvalues are integers. Nothing here quantizes anything: there is no Hilbert space, no commutation relation and no normal ordering. The integers come from the geometry of the commutator and not from a quantization rule.

No spin-statistics. The entry uses $(-1)^{2s}=+1$ as a statement about the action of a full turn on the carrier. That is a statement about the representation, and the connection between the sign and the statistics of a field is a separate theorem that the corpus does not prove here.

No statement about which particles exist. The carrier is a space, and the entry says what a spin-one field would look like if it took values there. It does not say that any particular boson is described by the triple, and the identification of a physical field with the carrier is always an extra step.

The list of three carriers is finite by the classification, not by survey. The entry cites the classification of the invariant irreducible subspaces and does not re-prove it. A reader who doubts that the list is complete has to read From the Spin to the Carriers and the Closures, where the three are derived and not enumerated.

Summary

The carrier. The vector part $\mathfrak{k}$ of real dimension three, the real pure elements, with $\vec a\vec b=-\vec a\cdot\vec b+\vec a\times\vec b$. It is closed under the commutator and not under the product, it has no axis and no family, and it is the only one of the three carriers that is a Lie algebra.

The action. $\operatorname{Ad}_U(X)=UXU^{*}$ at $N(U)=1$. A rotation preserves the triple and acts on it as an ordinary rotation, with residual $2.2\times10^{-16}$ over one hundred vectors. A boost does not: it sends the axis to $\cosh\phi\,\vec e_z-i\sinh\phi$ and the time line to $i\cosh\phi\,e_t-\sinh\phi\,\vec e_z$, both to $4.4\times10^{-16}$, so the image leaves the triple and stays in the sector.

The weights. $[\tfrac12\vec u,\vec v]=\vec u\times\vec v$ exactly; on the complexified carrier the weights are $+1$ on $\varepsilon_+$, $-1$ on $\varepsilon_-$ and $0$ on the axis; $2s+1=3=\dim_\mathbb{R}\mathfrak{k}$. The weight operator is $i$ times the commutator, Hermitian to $6.3\times10^{-16}$.

The full turn. $\exp(2\pi C(\vec u))=\mathrm{id}$ and $-e_t$ acts as the identity, to $8.9\times10^{-16}$, while a half turn sends $\vec u_2$ to $-\vec u_2$. So $(-1)^{2s}=+1$ and the double cover is invisible to the carrier.

The closure. $\mathbb{B}_\Phi=\mathbb{R}(ie_t)\oplus\mathfrak{k}$, of real dimension four, invariant under the whole group to $8.9\times10^{-16}$, a Lie algebra $\mathfrak{u}(1)\oplus\mathfrak{su}(2)$, and $1\oplus0$ under the rotations. The added direction is the central generator $ie_t$ and not the identity.

The two invariants. $N=\lvert\vec v\rvert^{2}-\lambda^{2}$, of the signs $(3,1)$, positive definite on the triple and negative definite on the line, the two summands orthogonal; $B^{2}=\lvert\vec v\rvert^{2}$, non-negative, with $\{B^{2}=0\}=\mathbb{R}(ie_t)$ on the sector.

The spectrum. $\{i(\lambda+\lvert\vec v\rvert),i(\lambda-\lvert\vec v\rvert)\}$ to $1.8\times10^{-15}$, purely imaginary, degenerate exactly on the added line. The cone is $\lvert\vec v\rvert=\lvert\lambda\rvert$ and meets the degenerate line only at the origin.

The reach. $\mathfrak{k}\setminus\{0\}\subseteq\Sigma_{11}$; the sector meets all four regions. The triple is the only carrier confined to one region, and the sector is the only object that spans the classification.

The two namings. The carrier of the spin is the triple, three dimensions, three states. The carrier of the amplitude is the sector, four dimensions, because a vector amplitude is a four-vector. The difference is one real dimension, which the Lorenz condition removes; the mass removes one more, and the counts run $4\to3\to2$.

The readings. Spin one is the adjoint representation, so the carrier is the algebra of the action. The three states are the triad of a frame. The integer weight is the price of two-sidedness and the boson does not pay the sign. A boost turns spin into time, so spin is not a four-vector. The rest frame is the degenerate locus. Mass is the norm and motion is the square of the vector part. Masslessness needs the added direction. The carrier is the algebra a gauge field takes values in. The field strength leaves it, and $\vec E$ and $\vec H$ sit in different halves. The sector has no complex structure, so the field is neutral. The constraint removes the singlet. The action is an automorphism only for rotations, and that is the little group.

Summary of Notation

symbol meaning
$\mathbb{B}$ the biquaternion algebra, real dimension $8$
$e_t,\vec e_x,\vec e_y,\vec e_z$ the four units, $e_t$ the identity, $\vec e_k^{2}=-e_t$
$i$ the central imaginary, commuting with the units
$\tilde Q=Q_te_t+\vec Q$ a general element, complex coefficients
$\mathfrak{k}$ the triple, $\mathbb{R}\vec e_x\oplus\mathbb{R}\vec e_y\oplus\mathbb{R}\vec e_z$
$\vec v$ an element of the triple, with $\lvert\vec v\rvert^{2}=v_x^{2}+v_y^{2}+v_z^{2}$
$\vec a\cdot\vec b,\ \vec a\times\vec b$ the scalar and vector products on the triple
${}^{*}$ Hermitian conjugation, $(Q_te_t+\vec Q)^{*}=\bar Q_te_t-\bar{\vec Q}$
${}^{\natural}$ natural conjugation, $\tilde Q^{\natural}=Q_te_t-\vec Q$
$\mathbb{B}_\Phi$ the material sector, $\{X^{*}=-X\}$, real dimension $4$
$\mathbb{B}_\Psi$ the informational sector, $\{X^{*}=X\}$, real dimension $4$
$N(\tilde Q)$ the norm, $Q_t^{2}+Q_x^{2}+Q_y^{2}+Q_z^{2}$
$B(\tilde Q)$ the polar form of the norm, with $B^{2}=Q_x^{2}+Q_y^{2}+Q_z^{2}$
$\Sigma_{ij}$ the region where $N$ is zero or not ($i$) and $B^{2}$ is zero or not ($j$)
$A(\vec u)=\tfrac12\vec u$ the rotation generator about the axis
$C(\vec u)=\vec u\times$ the adjoint operator of the generator on the triple
$\varepsilon_+,\varepsilon_-,\varepsilon_0$ the weight basis, circular and longitudinal
$\tilde S_3=i[\tfrac12\vec u,\ \cdot\ ]$ the Hermitian weight operator
$\tilde\Pi_\pm(\vec u)$ the idempotents $\tfrac12(e_t\pm i\vec u)$
$\operatorname{Ad}_U(X)=UXU^{*}$ the action of a Lorentz rotor of unit norm
$\tilde U_{\mathrm b}$ a boost rotor, $\cosh\tfrac{\phi}{2}e_t+i\sinh\tfrac{\phi}{2}\vec u_{\mathrm b}$
$\theta,\ \phi$ the rotation angle and the boost rapidity

Further Reading