The Biquaternion Basis of a Unified-Field Fibre Bundle
Introduction
The corpus meets the biquaternion algebra from inside its own programme: the algebra is chosen, the physical interpretations are then made, and every claim is a claim about what follows from the choice. This article records a source that arrives from the opposite direction. Its author is not working in the biquaternion programme and does not use the corpus's conventions; he asks what fermionic representations a unified field theory can carry, and reports that the basis of the principal fibre bundle underlying that theory is biquaternionic rather than complex. The paper is D. J. Cirilo-Lombardo, "Algebraic structures, physics and geometry from a Unified Field Theoretical framework", arXiv:1411.5493v3 [hep-th] (JINR Dubna, 2015).
The article has one job and it is deliberately narrow. Where the source's claim is a framework claim — the G-structure, the reduced tangent bundle, the incompleteness of Rothstein's theorem, the supermanifold construction — the article records the claim and says that it is a claim. Nothing here endorses the source's programme, and nothing here uses it as evidence for the corpus's: one paper that finds the algebra natural shows that the algebra can be found, not that it must be.
The source's central displays turn out to be the corpus's own objects, and finding them is the substance of the article. The structural group of the source's bundle is assembled from the two halves of one biquaternion built from a four-momentum; the source's demand that the group be inverted by its own conjugate transpose is the corpus's statement that this biquaternion has unit norm. Its Dirac equation is printed as a pair, and it is the corpus's one-sided pair. One of the source's claims, by contrast, does not reproduce: the metric that the source says is invariant under its group "due to its general form" is Euclidean and not Lorentzian, and the article records the discrepancy rather than repairing it.
The conventions are the corpus's and none is redefined. The algebra is $\mathbb{B}=\mathbb{C}\otimes_\mathbb{R}\mathbb{H}$, with basis $e_t=1,\vec e_x,\vec e_y,\vec e_z$ satisfying $\vec e_\eta^2=-e_t$ and $\vec e_x\vec e_y=\vec e_z$, and central imaginary unit $i$; the element is $\tilde Q=Q_te_t+\vec Q$ with complex coefficients $Q_\mu=q_\mu+iq'_\mu$. The norm is $\langle\tilde Q\rangle_{\natural}$, the sum of the squares of the four coefficients; the quaternion sign reverses the vector part, the star conjugates the coefficients and reverses the vector part, and the two agree on the scalar part. The two sectors are $\mathbb{B}_{\Psi}=\{\tilde Q:\tilde Q^{*}=\tilde Q\}$ and $\mathbb{B}_{\Phi}=\{\tilde Q:\tilde Q^{*}=-\tilde Q\}$. The algebra is $M_2(\mathbb{C})$ as a real algebra, and the ambient operator algebra is $M_4(\mathbb{C})$, whose real form $M_2(\mathbb{H})$ is what the source means by a quaternionic $2\times2$ algebra with the Dirac algebra as its complexification. These are the subjects of Algebra and Element Conventions, Norm and Pairing Conventions, The Informational Sector: States, Observables and Probability and The Material Sector: Four-Vectors, Interval and Lorentz Action, and nothing in them is restated here.
The body is written in the algebra. An operator is an element acting by multiplication, and a display — a matrix, a block, a column — is named only where the display itself is the object of the sentence, which in this article is wherever the source's equations are being reported. The two displays of one operator, and the one matrix that carries the body's to the source's, are Matrix Representations Conventions, §The Switch Between the Two Frames; the sixteen two-sided multiplications and the four frame basis operators are ibid., §The Sixteen Basis Operators in the Peirce Frame. The source's arrays are read in the display the corpus calls the Pauli one, which is the display of the matrix literature and of the mathematical corpus. The closing section prints, once, the arrays that the body does not.
The Source, Its Bundle, and Its Structural Group
The source works from a unified field theory it had proposed earlier, and its first structural object is a principal fibre bundle $P(G,M)$ over the real spacetime manifold $M$, whose structural group $G$ is meant to reflect the symmetries of the different fields. The paper's opening claim is that this bundle is the single geometrical structure of the theory — all matter and energy content comes from the same manifold — and its first technical claim, its Eq. (2), is that $G$ has a general form: a $4\times4$ array assembled from two $2\times2$ blocks $A$ and $B$, the first on the diagonal and the second off it with opposite signs, the two being the halves of one $2\times2$ array $Q$ — the scalar coefficient in $A$ and the three spatial ones in $B$.
The group is next written $G=A\,I_4+B\,J$, with $J$ the almost complex structure of the source's Eq. (3): the array that carries the identity in its off-diagonal blocks, with the two signs opposite. Its square is $-I_4$ exactly, so the source's printed relation $\sum_\lambda J_{\mu\lambda}J_{\lambda\nu}=\delta_{\mu\nu}$ needs a minus. The source calls the structure manifestly symplectic, and it exhibits three fundamental tensors built from $G$: $J$, a metric $g$ which it calls Lorentzian, and a third fundamental tensor, which the source prints with the letter sigma and two indices, with the relations between the three in its Eq. (5). It then imposes $G^+G=I_4$ — the condition whose meaning the next sections fix — and states the three compatibility conditions $\nabla_\lambda g_{\mu\nu}=0$, $\nabla_\lambda J_{\nu\lambda}=0$ and the vanishing divergence of the third tensor, in its Eqs. (14)–(16), with the remark that only two are independent.
Two verifiable details belong here as well, and both are slips against the source's own printed arrays: the square relation above needs the minus, and the timelike generator printed in Eq. (28) is not a member of an anticommuting set, as the Majorana section below shows.
In the corpus's reading neither object is unfamiliar. The almost complex structure is one two-sided multiplication: $J$ is $L_{e_t}R_{\vec e_y}$ in the source's frame, and its square is $-1$ because $\vec e_y^2=-\vec e_t$ — one line, and it fixes $J$ completely. The group is four basis operators: $G$ is the identity taken $a_0$ times plus the three spatial basis operators of the frame taken $a_k$ times, $G=a_0\,\mathrm{I}+\sum_\eta a_\eta\,iL_{\vec e_\eta}R_{\vec e_y}$, verified to zero for three generic complex sets of coefficients against the block array. The sum is one commuting basis operator and three that anticommute among themselves, and that is why the source's identities come out scalar: the mixed terms of any product of two such sums cancel, and only the squares survive.
Three of the later sections then draw the physical conclusions. The real Dirac structure of the manifold and the fermionic representations built on it are read algebraically (Sections II and IV); the Majorana representation and the non-compact structure behind the second sign pattern follow (Sections V–VI); the relation between the sign patterns and the dynamics, with the non-Hermitian CP operators, is drawn in Section VII; and the geometry is re-read as a G-structure on the tangent bundle $T(M)$, with the claim that the framework alters the spacetime structure rather than living on a fixed curved background (Sections VIII–IX, and the concluding list).
The corpus's interest is in that algebraic route and in the fact that the source takes it without the corpus's vocabulary. The source's own reason for expecting a biquaternion is stated in its Section I: the only generalized quaternion algebra over $\mathbb{C}$ is the ring of $2\times2$ matrices over $\mathbb{C}$. That is the corpus's own isomorphism $\mathbb{B}\cong M_2(\mathbb{C})$, and it is the one identification that the two programmes share verbatim.
The Dictionary: the Biquaternion Behind the Structural Group
The source writes its fermionic content in a $2\times2$ array with $a_0$ times the identity on the diagonal and its three-coefficient sum off the diagonal, the sum carrying the factor $-i$ that makes the array Hermitian; the four coefficients are determined physically in its Eqs. (19)–(21):
$$ a_0 = \frac{p_0}{m\,c}, \qquad a_\eta = i\,\frac{p_\eta}{m\,c}, \qquad \text{and} \qquad a_0^2 + a_1^2 + a_2^2 + a_3^2 = 1 . $$
The printed sum is consistent with the printed coefficients only because the $a_k$ are imaginary: with $a_k = i p_\eta/m$ the three squared terms are negative, and the condition reads $p_0^2 - \vec p^2 = m^2$, which is the source's own Eq. (21). In the corpus's letters the array is one element, and the source's four coefficients are its four coefficients in the basis $e_t,\vec e_x,\vec e_y,\vec e_z$:
$$ \boxed{\;\tilde Q = a_0 + \vec a\!\cdot\!\mathbf e,\qquad \langle\tilde Q\rangle_{\natural} = a_0^2 + a_1^2 + a_2^2 + a_3^2 = \frac{p_0^2-\vec p^{\,2}}{m^{2}c^{2}}\;} $$
The element is Hermitian, $\tilde Q^{*}=\tilde Q$: its scalar part is real and its vector part purely imaginary, which is what the source's display shows as a real diagonal and two conjugate off-diagonal entries. Two consequences follow, and together they are the identity the article is about. First, the norm of the element is the sum of the squares of its four coefficients, because the square of the vector part is a negative multiple of $e_t$; the identity is one line of the algebra, $\tilde Q\tilde Q^{\natural}=\langle\tilde Q\rangle_{\natural}e_t$, verified on three generic biquaternions to $10^{-16}$ and on the mass shell to $10^{-14}$. Second, the source's unit-norm condition of Eq. (19) is the mass shell of Eq. (21); the printed sum carries a minus in the three spatial terms only because the $a_k$ are imaginary. The two printed relations are therefore one identity, $\langle\tilde Q\rangle_{\natural}=1$.
The placement of $\tilde Q$ in the corpus's sectors is the part the source does not say and the corpus can. With $a_0$ real and the $a_k$ imaginary, $\tilde Q = a_0 + i\vec\beta\!\cdot\!\mathbf e$ with all four coefficients real, and that is a Hermitian element, $\tilde Q^{*}=\tilde Q$: the check is exact, with both sides equal and the real part of the vector part vanishing. So the source's structural object is an element of the informational sector $\mathbb{B}_{\Psi}$, and its norm is the norm of that sector. In the units $e_t,\vec e_x,\vec e_y,\vec e_z$, with the $i$ in the coordinate, $Q_t=q_t$ and $Q_\eta=iq_\eta$ for four real $q_A$, the norm reads
$$ \tilde Q\tilde Q^{\natural} = \left(q_t^2 - q_x^2 - q_y^2 - q_z^2\right)e_t , $$
of inertia $(1,3)$. The sign is worth naming, because it is the place where this article's conventions are easiest to misread: the coefficient of $e_t$ enters with a plus and the three spatial ones with a minus, because the $i$ sits in the spatial coordinates and is squared, $Q_\eta^2=(iq_\eta)^2=-q_\eta^2$; $\natural$ reverses the vector part, so only the squares survive. The mass shell is the level set $\langle\tilde Q\rangle_{\natural}=1$.
The corpus's own material four-momentum $\tilde P_{\Phi} = ip_0e_t + \vec p\!\cdot\!\mathbf e$ carries the same physics in the other sector, $\mathbb{B}_{\Phi}$, where the norm is negative; the exact dictionary between the two descriptions is one central factor and one conjugation,
$$ \tilde Q = \frac{-i\,\tilde P_{\Phi}^{\natural}}{m}, \qquad \langle\tilde P_{\Phi}\rangle_{\natural} = -m^2 , $$
so the plus of the informational normalisation and the minus of the material one are the same identity read from the two sides. This is the corpus's two-sided structure in its simplest instance: the source's fermionic object is the informational-side description, $\tilde Q$, and the material four-momentum is the other.
An interpretation the corpus declines. Because the printed $a_k$ are imaginary, the source's array is Hermitian, its scalar part real and its vector part purely imaginary. A reader who takes the printed sum $a_0^2+a_1^2+a_2^2+a_3^2=1$ with real $a_k$ obtains a different object, whose shell would be Euclidean, $p_0^2 + \vec p^2 = m^2$, and which the corpus's four-momentum interpretation excludes. The imaginary coefficients are not a typographical accident: they are what makes the printed relation match the printed relativistic relation. The point is recorded because the two interpretations coexist in the printed page and only one of them is the source's.
The Two Dirac Equations, the Two-Sided Operator, and Chirality
The source's Section IV prints a Dirac equation, and it prints it as two equations. Its Eqs. (26)–(27) are a pair of four-component equations for two objects, and its own biquaternion form, Eq. (31), is the same pair: one equation carries the biquaternion and the second carries the same biquaternion with its vector part reversed, which is the quaternion sign. The mass couples the two fields off the diagonal. The conjugation is an anti-automorphism, so it reverses the order of the multiplication, and the element form below records that reversal.
The source's two-component objects are the two halves of one element: the corpus's state is the element itself, and nothing is lost by writing the pair on the algebra. The position-space form of the same pair is
$$ \tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}, \qquad (\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}, $$
with both derivatives acting from the left. The first is the right-handed equation; the second carries $(\tilde\nabla_\Phi)^{\natural}$ because the left-handed field is the conjugate of the right-handed one. At $m=0$ the pair decouples into the two equations $\tilde\nabla_\Phi\tilde{Ξ}_{R}=0$ and $(\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=0$. Here $k_m=mc/\hbar$ is the mass wave number, the inverse of the mass length.
The two equations are one dynamics on the two halves, and the conventions own it: Operator and Calculus Conventions, §The Two First-Order Equations and Their Symbol, writes them as $\tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}$ and $(\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}$ and squares to $\Box_\Phi$. The corpus does not develop it here; what matters is which side each term acts on. The derivative acts on the left, the mass on the right, and the mass term carries each half into the other. That is the whole of the source's "mass couples the two fields off the diagonal", read on the algebra.
Chirality is the side, not an operator. The two halves are the $R$ and the $L$, and one basis operator of the frame separates them, $\mathsf{V}_{\mathrm{vol}}=-iL_{e_t}R_{\vec e_z}$ in the axis-free display, whose square is $+1$ and which is a pure right multiplier. So the split into the two halves is the split by that basis operator, and a projector would be the display of a split the pair already carries: the corpus names the half and does not compute the projector. Nothing in this article needs one, and nothing in it needs a matrix.
The generators printed in the source's Eq. (28) are not yet an anticommuting set; §The Majorana Transform: the Verified Effect and Two Printed Slips restores the timelike sign and shows that the repair leaves $G$ as displayed unchanged.
The Block Matrix, the Meaning of Its Conjugate Transpose, and the Invariant Form That Was Not Reproduced
The source's $G$ is then assembled from the same pair of blocks. With $A=a_0$ times the identity and $B$ the vector half, its Eqs. (3) and (4) give $G=A\,I_4+B\,J$, which is the four-basis operator sum recorded above. Both parts of the display were verified: $J^2 = -I_4$ exactly, so $J$ is an almost complex structure, and the sum reproduces $G$ exactly. The array commutes with $J$, so the block construction is complex-linear in the auxiliary unit, and with the physical coefficients the array is Hermitian, $G^\dagger = G$, with residual $0$.
The block form is not the entry-wise realification of the fermionic array. The shape of $G$ invites the interpretation that it is the real form of the complex array $Q$, in the way that a real array with two blocks in each row is the real form of a complex array split into its real and imaginary halves. The two are not the same object, and the difference decides what the source's plus means. The blocks are not the real and imaginary parts of $Q$: $A$ is the scalar half and $B$ the vector half, and they are Hermitian and anti-Hermitian, $A^\dagger = A$ and $B^\dagger = -B$, both with residual $0$, so $Q = A - iB$ is Hermitian. Replacing each entry of $Q$ by its real block returns $G$ only when $B$ is real, which with the physical coefficients means $a_1 = a_3 = 0$: at $p=(5,3,0,0)$ the two differ by $0.75$, while at $p=(5,0,3,0)$ they agree. The assembly is the realification of the pair $(A,B)$ in the auxiliary unit $J$, not of the entries of $Q$ in the central unit $i$. That is why $G^\dagger$ returns $G$ itself, while $G^+$ sends $B\mapsto -B$, which is the quaternion sign.
The source's unitarity condition is the next identity, and it is here that the corpus's dictionary fixes what the source's symbol means. The source's own computation of the diagonal of $G^+G$ is the sum of the squares of its two blocks; the display is
$$ G^+G = \langle\tilde Q\rangle_{\natural}\,I_4 , $$
with the off-diagonal blocks vanishing because $A$ is central, and $A^2+B^2 = \langle\tilde Q\rangle_{\natural}I_2$ because the three matrices anticommute. The identity holds with a worst deviation of $1.1\times10^{-16}$, and on the mass shell it is $G^+G = I_4$. On the element it is one line, $\tilde Q^{\natural}\tilde Q = \langle\tilde Q\rangle_{\natural}e_t$, and the same line says why: the time basis operator commutes and the three spatial ones anticommute among themselves, so the mixed terms cancel and what is left is the sum of the squares.
The same identity with the matrix adjoint fails, and the failure has a closed form: since $G^\dagger = G$ whenever $a_0$ is real and the $a_k$ are imaginary, $G^\dagger G = G^2$, and with $J^2 = -I_4$,
$$ G^\dagger G - I_4 = \big(\langle\tilde Q\rangle_{\natural}-1\big)I_4 + 2a_0\,J\,(B\oplus B) , $$
with $B\oplus B$ the array carrying two copies of $B$ on its diagonal, so the second term is the pair of blocks $B$ and $-B$ off the diagonal; the left side vanishes only for $B = 0$ and $a_0^2=1$; its entrywise maximum is $1.875$ at $p=(5,3,0,0)$, $m=4$. This is the corpus's distinction between the star and the quaternion sign, and the source's plus is the second. On the element the two candidate conditions are $\tilde Q^{\natural}\tilde Q = \langle\tilde Q\rangle_{\natural}e_t$ and $\tilde Q^{*}\tilde Q = \tilde Q^2$: the first is the norm, the second is the square, and the square is the norm only for $\tilde Q=\pm e_t$, because $(\tilde Q)^2-\langle\tilde Q\rangle_{\natural}e_t = 2\vec q^2e_t+2a_0\vec a\!\cdot\!\mathbf e$ does not vanish otherwise. The interpretation is confirmed by computation and not assumed, because the two readings agree on every array whose two halves are real, and the source's coefficients are not real.
A claim that did not reproduce, and the computation that decides it. The source asserts, after its Eq. (2), that there exists a Lorentzian metric $g_{\lambda\mu}$ invariant under $G$ "due to its general form (2)"; the inertia it attaches to that metric is printed in its own spelling in §The Legacy Form of the Source's Metric Claim. The paper's own equations defeat the sentence. With $A = a_0$ times the identity and $B$ the vector half, direct multiplication gives the transpose product whose off-diagonal block is $a_0(B - B^{\mathsf T})$, and $B$ is symmetric exactly when the coefficient of the one antisymmetric of the source's three matrices vanishes. On that slice the transpose product is $\langle\tilde Q\rangle_{\natural}I_4$ exactly — the source's own unit-norm condition — and the source's $G^+$ is the ordinary transpose there, so Eq. (2)'s $G^+G = I_4$ reads literally as $G^{\mathsf T}\delta\,G = \delta$. The symmetric form Eq. (2) fixes is therefore the Euclidean $\delta$, of inertia $(4,0)$, and not a Lorentzian one. $G$ also commutes with $J$, and on the same slice it preserves $J$ up to the norm,
$$ G^{\mathsf T}JG = \langle\tilde Q\rangle_{\natural}\,J , $$
which is $J$ itself on the unit-norm slice. Both identities were verified exactly at six rational on-shell points with the antisymmetric coefficient zero, the physical point $p = (5,3,0,0)$, $m = 4$, among them; off the slice neither holds. On the mass shell the two residuals have the same Frobenius norm, because $G$ commutes with $J$ gives $G^{\mathsf T}JG-J=(G^{\mathsf T}G-I)J$ and $J$ is orthogonal.
The transpose is not a conjugation of the element, and the article does not force it into one. On the basis operators of the display the transpose exchanges the two slots, so a transpose product is a product of the two sides and not a norm; the fact proved above is a fact about the display, which is the level at which the source's sentence was written.
The whole space of invariant symmetric forms can then be computed, and it contains no Lorentzian element. On the slice, the invariant forms solve a linear system whose solution space is four-dimensional, and the common commutant of that space is two-dimensional and contains a complex structure $K$, $K^2=-I_4$. Every invariant form therefore commutes with $K$, so its eigenvalues are doubled, and its inertia is $(4,0)$, $(0,4)$ or $(2,2)$ — never $(1,3)$, never $(3,1)$.
Solved exactly in rational arithmetic at six on-shell points with the antisymmetric coefficient zero, the space is four-dimensional; over a hundred integer combinations of its basis elements the inertias realised were the three nondegenerate ones above, together with the degenerate ones from combinations that drop rank; and for the group generated by three of those $G$ the invariant symmetric form is one-dimensional and Euclidean.
Where the antisymmetric coefficient does not vanish the space changes — four-dimensional with inertias $(2,2)$ and the degenerate $(1,1)$ at the physical point $p=(5,0,3,0)$, $m=4$ and at a real point with $a_2 = 4/5$, and empty at the physical point $p=(0,5,3,0)$, $m=4$ — and no Lorentzian form appears in any of them either.
The sentence therefore cannot be repaired by choosing another metric, and Eq. (7)'s $G \subset L(4)$ is where the discrepancy enters: the displayed $G$ is compact — $*$-unitary — and not Lorentz-orthogonal. The citation the sentence carries, the source's [31] (Bishop and Goldberg, Tensor Analysis on Manifolds, ch. 5 p. 208), is given "about the meaning of Lorentz metric" — a definition, not a derivation from Eqs. (2)–(4).
The earlier, weaker check is subsumed by this one. Over sixty random $G$ built from real unit coefficients on the sphere, and at the physical point built from the imaginary coefficients, every candidate form tried failed $G^{\mathsf T}g\,G = g$, with residuals between $3$ and $4.4$ at the physical point; only the unit metric passed, and only on the slice the computation above isolates. The candidates are legacy, and they are written in the source's own arrays in §The Legacy Form of the Source's Metric Claim. That is why a search over a fixed list of candidates could not settle the sentence: the list did not contain the form the computation would find.
The Majorana Transform: the Verified Effect and Two Printed Slips
The source's Section V performs Majorana's 1937 transformation on the same $G$, with the aim stated in its own words: to obtain a $*$-unitary transformation for which the Dirac equation becomes "with real coefficients and symmetric for both: fermions and antifermions". The transformation is a $4\times4$ array built from the one antisymmetric of the source's three matrices, and it is attributed by the source to Majorana. Basis operator for basis operator it is a sum of two basis operators, $\tfrac{1}{\sqrt2}\bigl(iL_{e_t}R_{\vec e_z}-L_{\vec e_y}R_{\vec e_x}\bigr)$, so written as an action on the algebra it is $\tilde Q\mapsto\tfrac{1}{\sqrt2}\bigl(i\tilde Q\vec e_z-\vec e_y\tilde Q\vec e_x\bigr)$. It is not a conjugation by an element: it is a sum of two basis operators, and its square is the identity only because the cross terms of the two cancel — which they do, since $\vec e_x\vec e_z=-\vec e_y$. It was verified to be involutive and $*$-unitary, its square and its product with its adjoint both the identity array, with a worst deviation of $2.2\times10^{-16}$. It is in fact its own adjoint, so it is its own inverse.
The source's own generators are printed in its Eq. (28): the identity in the timelike slot and, in the three spatial ones, the three basis operators with the opposite sign. As printed they are not an anticommuting set: the anticommutator of the timelike generator with each spatial one is twice that spatial generator, which does not vanish, because the printed timelike array is the identity and the identity commutes with every basis operator. So the displayed timelike generator has lost a sign and must be the negative of the fifth basis operator, $iL_{e_t}R_{\vec e_z}$. With that single correction the four generators satisfy the whole algebra with the weight $\varepsilon=(1,-1,-1,-1)$, $\{\gamma_\mu,\gamma_\nu\}=2\varepsilon_\mu\delta_{\mu\nu}I_4$, and their squares form $\varepsilon_\mu I_4=\mathrm{diag}(+1,-1,-1,-1)$: every square and every anticommutator verified exactly. The correction repairs the Dirac frame of Eq. (28) and leaves $G$ itself untouched, because $G$ is written block by block with the scalar half on its diagonal.
The shorthand and the display use different conventions. Before its Eq. (35) the source writes $G$ as $a_0$ times the identity plus its three spatial generators weighted by $a_k$, and adds that the gamma is "in the standard form". That caveat is doing real work, because the two conventions do not agree. With the standard generators — the negative of the fifth basis operator in the timelike slot and the three spatial basis operators with a plus — the sum is exactly $G$, verified to $0$. Eq. (28) prints the identity in the timelike slot and the three spatial entries with a minus, and with those the sum misses $G$ by twice its vector part. So the letters denote one thing in Eq. (28) and another in the shorthand, and the convention that reproduces Eq. (2) is the shorthand's.
The verified effect. The source's Eq. (34) is the transformation the corpus needs, and with it the source's stated purpose is achieved. Recomputing the conjugated group term by term and recomposing it gives
$$ G_2 = a_0\,\mathrm{I} - a_1\,L_{\vec e_z}R_{e_t} - a_2\,iL_{\vec e_y}R_{\vec e_y} + a_3\,L_{\vec e_x}R_{e_t} , $$
with a residual of $4.4\times10^{-16}$, so the three printed transforms reproduce from the printed transformation. The third is the printed one verbatim; the two diagonal ones are the printed ones with the two factors of the product exchanged. Each transformed generator is purely imaginary, and because the source's own $a_k$ are purely imaginary the combination weighted by them is real. The recomputed $G_2$ is therefore real and symmetric — the imaginary part vanishes identically and $G_2 - G_2^{\mathsf T} = 0$. That is the source's stated purpose, and it is confirmed.
Two printed slips. The diagonal half of the printed $G_2$ in the source's Eq. (38) carries the sum of its two off-diagonal products; the recomputation gives their difference, the two differing by twice $a_3$, that is by $2|a_3|$ in the units of the coefficients ($1.5$ at $p=(5,0,0,3)$, $m=4$). And the printed sentence that $G_2$ and its transpose satisfy $G_2G_2^{\mathsf T} = G_2^{\mathsf T}G_2 = I_4$ does not reproduce: because the recomputed $G_2$ is real symmetric, both products are $G_2^2$, and since the transformation is unitary,
$$ G_2^2 - I_4 = \mathsf{U}\bigl(G^2 - I_4\bigr)\mathsf{U}^{-1} , $$
which is the dagger failure of the previous section in a transformed frame and is not small: its Frobenius norm is $4.37$, the same as that of $G^2-I_4$, at either on-shell point tested. The identity the corpus verified is the one with the sign ${}^{\natural}$, before the transformation, $G^+G = \langle\tilde Q\rangle_{\natural}I_4 = I_4$ on the mass shell; the corpus records the post-transformation sentence as not reproduced and does not repair it.
The Non-Compact Triple, the (2+2) Inertia, and the Two Real Forms
The source's Section VI introduces a second triple of $2\times2$ arrays, and this one the corpus can check completely. The three are the source's first matrix, its third, and its second with a factor $i$ taken out; the middle one is the source's non-compact element. In the corpus's letters the triple is
$$ i\vec e_x,\qquad \vec e_y,\qquad i\vec e_z , $$
so the middle element is the only one whose square is $-e_t$, the other two being square roots of $+e_t$. The products are the quaternion products read with the two imaginary units,
$$ i\vec e_x\cdot\vec e_y=i\vec e_z,\qquad \vec e_y\cdot i\vec e_z=i\vec e_x,\qquad i\vec e_z\cdot i\vec e_x=-\vec e_y , $$
which are the source's three printed products, including the minus in the third, all exact.
These are the generators of the split real form of the triple the corpus meets as the rotation generators. The compact triple of imaginary units satisfies $[i\vec e_x,i\vec e_y]=-2\vec e_z$; the source's satisfies $[i\vec e_x,\vec e_y]=2i\vec e_z$. The two differ by one factor of $i$, and the $i$ is exactly what the third element differs by: two of the source's three elements are imaginary units and one is a vector unit, so a product of two carries $i$ or $i^2$, and the relation comes out with the structure constant of the quaternion product and no further factor. The middle element carries the source's non-compactness and the paper says so: its two outer elements are imaginary units and its middle one a vector unit, so the squares are $\vec e_y^2=-e_t$ against $(i\vec e_x)^2=(i\vec e_z)^2=+e_t$, all three verified.
The triple is the algebraic content of the source's other sign pattern, and the source shows the mechanism rather than asserting it. Its Eq. (43) repeats the unit-norm condition with the sign of the second coefficient reversed,
$$ \boxed{\;a_0^2 + a_1^2 - a_2^2 + a_3^2 = 1\;}, $$
and the paper attributes the change to the square of its non-compact element. In the corpus's letters the mechanism is exactly the squares. The coefficient $a_2$ multiplies the one element of the second triple whose square is $-e_t$, so the element the second triple carries has the plain norm
$$ \langle a_0 + a_1i\vec e_x + a_2\vec e_y + a_3i\vec e_z\rangle = a_0^2 + a_1^2 - a_2^2 + a_3^2 , $$
the quaternionic norm of the same element being the undifferentiated $a_0^2+a_1^2+a_2^2+a_3^2$; the two forms differ by the sign of the one direction whose square is negative, and no coefficient is relabelled. With the same coefficients $a_0 = p_0/(m\,c)$ and $a_k = ip_\eta/(m\,c)$ this is the $(2{+}2)$ relation
$$ p_0^2 - p_1^2 + p_2^2 - p_3^2 = m^2 , $$
verified exactly at $p=(5,3,0,0)$, $m=4$, against the $(1,3)$ relation of Eq. (21). The source itself then says that "the physical role of the coefficients $a$ cannot be easily identified as before" and lists two possibilities; the corpus records both the mechanism and the admission.
The paper's wider claim is that the non-compact symmetry of what it calls a hidden (bi)quaternionic structure becomes visible in $(2{+}2)$ — two timelike and two spacelike directions — and it lists convergence with work on $(2{+}2)$ physics, two-dimensional black holes and $N=2$ strings. The corpus's interpretation is that this is a genuine second real form of the same algebra and not a relabelling of the first: the split real form has its own group of invertibles and its own polar decomposition, which the corpus treats in the split-quaternion articles, and $(2{+}2)$ is the inertia of that form just as $(1,3)$ is the inertia of the material sector's norm.
Reality, the Non-Hermitian Operator, and the Corpus's Sign Classification
Two of the source's Section VII claims bear directly on articles the corpus already owns.
The first is that the availability of a Majorana representation is linked to the spacetime signs. The corpus agrees, and has said it in more detail: in Real Spinors and Reality Conditions on the Biquaternion Algebra with Hermitian Adjoint the reality type of the module is a function of the inertia of the quadratic form modulo eight, and in that classification the corpus's ambient module admits a real structure in inertia $(1,3)$ while the internal space is of complex type and admits none. The source states the dependence and gives no table; the corpus's contribution is the table, and the source's Section V is a worked instance of it.
The second is the claim that the framework produces non-Hermitian (CP) dynamic operators, and that this is structural rather than exotic. Here the corpus's translation separates two things the source's prose runs together. The structural object $\tilde Q$ is Hermitian: with the source's own coefficients, $\tilde Q\in\mathbb{B}_{\Psi}$, checked with residual $0$. The non-Hermiticity the source flags belongs to the generalized momentum built in its Eqs. (54)–(56), where the paper notes that in the $3{+}1$ case the quantity $E^2 + H^2$ can be negative "due to the non-hermitian character of the generalized momentum operators". So the corpus's interpretation is that the non-Hermiticity lives in the kinetic operator and not in the biquaternion the bundle's basis carries. The distinction matters because the corpus's framework is Hermitian throughout — its Born rule is the Hermitian form of a Hermitian element, read as the Hermitian pairing — and a genuinely non-Hermitian dynamical generator would be a structural departure of a different size from a change of signs. The source asserts it as a consequence of the $G$-structure; the paper contains no worked dynamics in which it is exhibited, and the corpus records it as an assertion.
The Bundle Claim: Which Naturality Is Meant?
The paper's central algebraic claim is that the basis of $P(G,M)$ carries a biquaternionic structure rather than a complex one. The corpus's existing bundle-theoretic material reaches the same algebra from the other end, and the difference is worth stating.
The Hopf Fibration and the Biquaternion Gauge Bundle constructs a principal bundle whose structure group is a subgroup of the algebra's group of invertibles, with the Hopf map as the projection and the defining module as the associated vector bundle. That construction is pointwise: the algebra acts on the fibre, and the bundle's local data are the transition functions of the frame. The source's claim is about the fibre algebra itself — that the basis in which the fibres are expressed is biquaternionic — and it is a stronger statement, because a basis is the thing the corpus's Curved Spacetime and the Biquaternion Framework showed cannot carry curvature.
The corpus declines to accept the word naturally without specifying which naturality is meant, and the article states the two candidates. A preferred basis claim would say that the bundle admits a reduction of its structure group to a subgroup of the biquaternion group of invertibles; that is a statement about $G$ and can be tested, and the source's own test is the unitarity condition $G^+G = I_4$ — which, as the previous sections show, holds in the $\natural$-sign interpretation and not in the ordinary one. An algebra-on-the-fibres claim would say that each fibre is a module over $\mathbb{B}$; that is what the corpus's Hopf article already constructs, and it is a statement about the fibres rather than about the basis. The two are not the same, and the paper's wording covers both. The corpus's position is that the second is true and unremarkable, that the first is the substantive claim, and that the verification of $G^+G = I_4$ in the $\natural$-sign interpretation is what makes the first operative.
The G-Structure, the Reduced Tangent Bundle, and Rothstein
The source's concluding list makes a series of claims about geometry, and the corpus records them as claims, with the articles that would have to carry them named.
A consistent unified theory, the source says, must rest on a G-structure for the tangent bundle $T(M)$ reflecting the symmetries of the fields considered. The corpus's Curved Spacetime and the Biquaternion Framework reached a related conclusion from the other side: the algebra supplies the local Lorentz group at every point and can even supply the connection and its curvature as elements of its own Lie subspace, but it supplies nothing that determines them, so the content of a biquaternion theory of gravitation must lie in the field that attaches the algebra to the manifold. A G-structure is precisely such a field, and the corpus's The Einstein Field Equations in the Biquaternion Framework: The Linearised Dictionary and the Non-Linear Gap is where a structure-group mechanism would be recorded. The source does state the three compatibility conditions its own tensors must satisfy, Eqs. (14)–(16); what it does not give is the reduction itself, nor a curvature computed from it. It states the requirement, and the requirement is one the corpus has already identified independently.
The second claim is the sharpest, and the corpus notes it as the source's most distinctive sentence: whereas standard quantum field theory in curved spacetime leaves the spacetime structure alone — its structure group remains Lorentzian — this framework alters the structure, because the structure group of the reduced tangent bundle becomes that of the induced field theory, with the same curvature of the tangent space. That is a mechanism by which field content changes the geometry rather than riding on it, and it is different in kind from the corpus's tetrad route, in which the frame carries a background metric. The corpus records the claim as unverified: the paper gives no reduction and no curvature computation for it.
The third and fourth claims are that the spacetime signs and non-Hermitian dynamic operators are both induced by the same G-structure, and that torsion, through its dual four-vector, plays a key role in both the signs and the CP character of the dynamics. The corpus has torsion material in Curved Spacetime and the Biquaternion Framework, where metric compatibility and vanishing torsion are conditions to be imposed and not consequences of the algebra; the source's claim is that a specific torsion field, the dual of a totally antisymmetric torsion, appears as a consequence of the third fundamental tensor of the tangent space. The two statements are compatible and neither implies the other: the source's third tensor is a further structure the corpus does not have.
Finally, the source claims that Rothstein's theorem is incomplete for its description of the spacetime manifold, and that the Darboux–Kostant theorem is satisfied so that $M$ fits a general supermanifold. The corpus has no Rothstein material and no supermanifold material, so it neither confirms nor disputes the claim; it is recorded because the source presents it as a result of the same construction, and because the corpus's own parity article is the nearest thing it has to a boson–fermion grading of the manifold.
The Source's Interlude on the Algebra of the Frame
The source's Section III is a short interlude arguing that the natural language of the description is an algebra of the frame and not arrays. The argument has two parts and the second is the interesting one. In the orthogonal kind of algebra, the source says, the symmetric product of two basis vectors is the inner product and the antisymmetric product is a bivector; in the symplectic kind the roles are exchanged, the antisymmetric product giving the metric and the symmetric one a bivector. The generators of the orthogonal kind can be brought to a basis in which they act as creators and annihilators of the fermionic kind, and those of the symplectic kind as the bosonic ones, so the two kinds can be united into one structure whose basis spans what the source calls a superspace.
The corpus carries the first half of this in its own terms and not the second. The two products the source describes are the corpus's own two: the symmetric product $\tilde Q_1\bullet\tilde Q_2$ carries the form in its scalar part and the wedge $\tilde Q_1\wedge\tilde Q_2$ carries the bivector (Product and Form Conventions, §The Twelve Products and Their Names), so the exchange of roles that separates the source's two kinds is the exchange of those two parts, and the corpus's four general products carry the four readings. What the corpus does not carry is the doubling: the source places two algebras of different type side by side and unites them, and the corpus's articles contain one algebra with its two sectors and its grading. Biquaternion Versors and the Orthogonal Group and the parity article do give the grading the fermionic half needs — the even part is the algebra and the odd part is the fermionic slot, so that two fermionic steps compose to a bosonic one — but that is a grading of one algebra and not the doubling. The interlude is recorded as the source's proposal, and the note it draws from it, that the manifold's own structure supports a consistent quantum theory with a clear geometrical meaning, is recorded with it and not endorsed.
A Constructed Biquaternionic Basis, from an Independent Source
The bundle-basis idea recorded above is a claim about a basis, and it is worth setting beside it an independent source in which a biquaternionic basis is actually built and used. J. Fredsted, Spinor fields without Lorentz frames in curved spacetime using complexified quaternions (arXiv:0811.1357v4 [math-ph], 2009), abandons the manifold frame — "thus leaving the arena of Riemannian manifolds" — and takes as primitive a basis $s_\mu \in \mathbb{B}_{\Phi}$ of the biquaternion algebra itself, with the metric given by the algebra's bilinear form,
$$ g_{\mu\nu} = \langle s_\mu, s_\nu\rangle_{\natural} , $$
whose nondegeneracy follows from the linear independence of the four basis elements. Local Lorentz freedom is the conjugation $s_\mu \mapsto \Lambda s_\mu \Lambda^{-1}$ by an element of unit norm — the source prints the inverse with its own bar — and a single algebra-valued connection $\omega_\mu \in \mathbb{C}\otimes\mathbb{H}$, subject only to $\mathrm{Sc}(\omega_\mu + \omega_\mu^{*}) = 0$, serves both the tensor and the fermionic fields; its imaginary-scalar part is a local $\mathrm{U}(1)$ freedom and its vector part is the Lorentz connection. The corpus's Curved Spacetime and the Biquaternion Framework records the same construction from its own side, and the two are the same structure in different notation: the basis is this article's material-sector frame field, the metric is the quaternionic form the corpus already computes, and the connection is the two-sided gravitational connection of the corpus's material sector.
The two external instances sit in opposite sectors, and that is the point of recording both. The source of this article builds its structural group from a biquaternion that is Hermitian of unit norm — an element of $\mathbb{B}_{\Psi}$, the informational sector — and its claimed invariant Lorentzian metric does not reproduce. Fredsted's basis is anti-Hermitian, $s_\mu \in \mathbb{B}_{\Phi}$, the material sector, and his metric is real and Lorentzian by construction, its invariance under the conjugation being the identity the corpus's frame route already verified. So the biquaternionic-basis proposal has, in the material sector, a working instance in which the metric comes out of the algebra rather than being asserted of it.
The instance does not build $P(G,M)$, and it does not supply dynamics. Fredsted's basis is a basis of the algebra over the manifold, not the basis of a principal bundle with a structural group; and there is no action for $s_\mu$ or for $\omega_\mu$, no field equation, and no curvature computed from the connection. The construction therefore raises the floor of the bundle-basis idea without touching the ceiling this article's source leaves in place: a biquaternionic basis, whether claimed or constructed, is a pointwise structure until a field equation attaches it to the manifold.
What the Source Shows, and What It Does Not
The value of the paper to the corpus is a single fact, and the article states it plainly. A mainstream high-energy paper, published without any connection to the biquaternion programme, asks what algebraic structure the basis of its bundle should carry and answers with the biquaternion algebra — the algebra that the corpus's A Brief History of Biquaternions in Physics records as a nineteenth-century construction and that the corpus's The Empirical Status of the Biquaternion Framework treats as a framework with a proposal attached. That is external evidence that the algebra can appear unsought, and it is the kind of evidence the corpus's empirical-status article is organised to weigh. It is not evidence that the algebra must appear, and the source's own programme is not thereby supported: the paper's framework claims are exactly as unverified here as they would be if written in any other notation.
Two caveats belong with the record. The first is the word QFT: in the source's conclusion it denotes the author's own unified construction, not quantum field theory as the expression is normally used, and the corpus disambiguates before quoting. The second is the extraction itself: the automated summary of the paper carries fragments of the source's matrices, with their layout lost, and the transformed generators of Eqs. (35)–(37) come through as fragments of products. The corpus quotes those three only after the recomputation has fixed them — the two diagonal ones agree with the recomputed basis operators up to the order of the factors — and quotes no other entry of the paper.
The Legacy Form of the Source's Metric Claim
A sign tuple is legacy: it is a tuple of four signs, and the corpus reads its intervals on the coordinate, where every sign is the square of an $i$. The reason is in Frame and Metric Conventions, where the interval is read on $Q_\mu$ and the tuple is a coefficient tuple and not a sign pattern. The tuples are written here once, in the source's own spelling, so that a reader who opens the paper recognises the sentences the body answers. Nothing above depends on them.
The claim. After its Eq. (2) the source asserts a Lorentzian metric $g_{\lambda\mu}$ of inertia $(+,-,-,-)$ invariant under $G$ "due to its general form (2)". §The Block Matrix, the Meaning of Its Conjugate Transpose, and the Invariant Form That Was Not Reproduced answers it: on the slice where $B$ is symmetric the source's $G^+$ is the ordinary transpose, its condition reads $G^{\mathsf T}\delta\,G=\delta$, the form it fixes is the unit metric $\delta_{\mu\nu}=(1,1,1,1)$, and the space of invariant symmetric forms contains no element of inertia $(1,3)$.
The candidate list, as it was tried. $\mathrm{diag}(+1,-1,-1,-1)$, $\mathrm{diag}(-1,+1,+1,+1)$, $\mathrm{diag}(+1,+1,-1,-1)$ and its negative, and the split form with the identity in the off-diagonal blocks, all failed $G^{\mathsf T}g\,G=g$ — over sixty random $G$ built from real unit coefficients on the sphere, and at the physical point built from the imaginary coefficients, with residuals between $3$ and $4.4$. Only the unit metric passed, and only on the slice the computation isolates. The arrays are kept for a reader who wants to repeat the search; the body states the outcome and uses none of them.
The Legacy Pauli Gamma Formulation
The Pauli and gamma formulation is legacy: the reasons are explained in the article The AXF Matrices: Axis-Free, Integer Entries, §Why the Pauli and the Gamma Displays Are Refused. The generators carry no frame label: $\gamma_\mu$ is one set of matrices, and the frame is named on the coordinate.
The gammas are a frame of $\mathrm{Cl}_{1,3}(\mathbb{C})=M_4(\mathbb{C})$, not a different algebra. This is the only section of the article in which they appear; every operator statement above is written in the algebra or in the basis operators it names.
Here $a,b\in\mathbb{B}$, $L_a$ and $R_b$ are the left and right multiplications, $L_a\tilde{Ξ}=a\tilde{Ξ}$ and $R_b\tilde{Ξ}=\tilde{Ξ} b$, and $c(M)$ is the coefficient matrix of an element $M$ in the basis of the sixteen products.
$$ \begin{aligned} \gamma_0 &= iL_{\mathrm{pau},e_t}R_{\mathrm{pau},\vec e_x}, &\gamma_1 &= iL_{\mathrm{pau},\vec e_x}R_{\mathrm{pau},\vec e_y},\\[2pt] \gamma_2 &= iL_{\mathrm{pau},\vec e_y}R_{\mathrm{pau},\vec e_y}, &\gamma_3 &= iL_{\mathrm{pau},\vec e_z}R_{\mathrm{pau},\vec e_y},\\[2pt] \gamma_5 &= -iL_{\mathrm{pau},e_t}R_{\mathrm{pau},\vec e_z}. \end{aligned} $$
The same in the axis-free display:
$$ \begin{aligned} \gamma_0 &= iL_{\mathrm{axf},e_t}R_{\mathrm{axf},\vec e_x}, &\gamma_1 &= iL_{\mathrm{axf},\vec e_x}R_{\mathrm{axf},\vec e_y},\\[2pt] \gamma_2 &= iL_{\mathrm{axf},\vec e_y}R_{\mathrm{axf},\vec e_y}, &\gamma_3 &= iL_{\mathrm{axf},\vec e_z}R_{\mathrm{axf},\vec e_y},\\[2pt] \gamma_5 &= -iL_{\mathrm{axf},e_t}R_{\mathrm{axf},\vec e_z}. \end{aligned} $$
Each generator carries one unit $i$, and a product of two carries $i^2=-1$; that sign is where the weight enters, $\{\gamma_\mu,\gamma_\nu\}=2\varepsilon_\mu\delta_{\mu\nu}I_4$ with $\varepsilon=(1,-1,-1,-1)$.
The five are generators, not a basis: the basis is the sixteen products. A single product is a pure tensor; the pure tensors form the seven-dimensional cone $\operatorname{rank}c(M)\le1$, and a general element is a sum of at most four, with $\ell(M)=\operatorname{rank}c(M)$.
The first four are the frame basis operators $C_\mu$ of the body, $C_t=iR_{\vec e_x}$ and $C_\eta=iL_{\vec e_\eta}R_{\vec e_y}$, and $\gamma_5$ is the fifth basis operator $\mathsf{V}_{\mathrm{vol}}=-iL_{\mathrm{pau},e_t}R_{\mathrm{pau},\vec e_z}$; the gamma names are the Dirac-algebra names for the same basis operators. The central grading $\chi=i\tilde{Ζ}_{\mathrm{ad}}$ — the difference of the two central idempotents of the complexified algebra, so $\chi^2=e_t$ — is a different object, represented by $\gamma_5$ on the Dirac module.
The dictionary is complex-linear; over $\mathbb{R}$ the products span $\mathrm{Cl}_{1,3}(\mathbb{R})\cong M_2(\mathbb{H})$ and the basis operators span $\mathrm{Cl}_{3,1}(\mathbb{R})\cong M_4(\mathbb{R})$.
The display. Every array printed in this section is given twice: in the Peirce display first, labelled $\mathrm{pau}$, whose arrays are the ones the literature prints, and in the axis-free display after it, labelled $\mathrm{axf}$, which is the display of the body and of the corpus. The two are carried into one another by $L_{\mathrm{axf},e_\mu}=C_{\Pi}L_{\mathrm{pau},e_\mu}C_{\Pi}^{-1}$, and a conjugation by $C_{\Pi}$ changes no product, no anticommutator, no trace and no determinant, so every identity of the section holds in both. An array is a statement about one display; a basis operator named in the running text keeps the Peirce label, and its axis-free array is the conjugate. The label names the array and not the operator.
Summary
The source is a mainstream high-energy paper whose central algebraic claim is that the basis of the principal fibre bundle of a unified field theory carries a biquaternionic structure rather than a complex one. The article translates the algebra it displays into the corpus's dictionary and checks what can be checked: the translation is exact on every display the source prints consistently, and the displays it does not print consistently are named.
The structural group of the source's bundle is assembled from the two halves of the biquaternion $\tilde Q$ built from a four-momentum by $a_0 = p_0/(m\,c)$ and $a_k = ip_\eta/(m\,c)$. The source's unit-norm condition $a_0^2+a_1^2+a_2^2+a_3^2 = 1$ and the source's relativistic relation $p_0^2-\vec p^{\,2} = m^{2}c^{2}$ are one identity, $\langle\tilde Q\rangle_{\natural} = 1$, verified to $10^{-14}$; with the imaginary coefficients the printed sum carries its signs, so the source's object is the corpus's Hermitian biquaternion of unit norm — an element of the informational sector $\mathbb{B}_{\Psi}$, in the corpus's normalisation the opposite of the material four-momentum's $\langle\tilde P_{\Phi}\rangle_{\natural} = -m^{2}c^{2}$. The source's $G^+G = I_4$ holds with the quaternion sign and fails with the matrix adjoint, which fixes the meaning of its symbol; the block assembly is the realification of the pair of halves in the auxiliary unit $J$, not the entry-wise realification of the fermionic array, and the two agree only when the vector half is real. The group is four basis operators — one that commutes and three that anticommute among themselves — which is why its identities come out scalar.
The Majorana transformation the source uses is $*$-unitary and involutive, and it achieves its stated purpose: the conjugated group is real and symmetric, because the transformed generators are purely imaginary while the source's own coefficients $a_k$ are purely imaginary. The source's Dirac equation is a pair, its Eqs. (26)–(27), and the corpus reads it as the one-sided pair $\tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}$, $(\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}$, which is one operator on the two halves, with the mass carrying each into the other and the fifth basis operator separating them. Two printed slips are recorded: the printed generators of Eq. (28) are not an anticommuting set until a sign is restored in the printed timelike generator, and the diagonal half of the printed transformed group carries the wrong sign in its two off-diagonal products; the shorthand the source uses for the group before its Eq. (35) is moreover correct only for the standard generators, which are minus the printed spatial entries. The non-compact triple is the source's split real form, it reproduces its three printed products exactly, and it is the triple $i\vec e_x$, $\vec e_y$, $i\vec e_z$; the source's own coefficient sum in that sector then reads the $(2{+}2)$ relation $p_0^2-p_1^2+p_2^2-p_3^2=m^{2}c^{2}$, which is the plain norm of the element that triple carries.
One claim does not reproduce, and the paper's own equations are what defeat it: the Lorentzian metric the source says is invariant under the group is not invariant under it, because Eq. (2)'s condition $G^+G = I_4$ is the Euclidean orthogonality condition whenever the coefficient of the one antisymmetric matrix vanishes, and the whole space of symmetric forms invariant under the group contains no form of inertia $(1,3)$. The group is compact — $*$-unitary — rather than Lorentz-orthogonal, and Eq. (7)'s $G \subset L(4)$ is where the discrepancy enters. The framework claims — the G-structure, the reduced tangent bundle with the induced field theory's structure group, the torsion-induced sign and CP character, the incompleteness of Rothstein's theorem — are recorded as claims, with the corpus's own curved-spacetime and gravity articles named as the places where the geometry would be built. The source's contribution to the corpus is the external instance, not the programme: the biquaternion algebra appears here unsought, in a paper that shares none of the corpus's interpretations, and that is worth recording without being mistaken for confirmation.
Summary of Notation
| Symbol | Meaning |
|---|---|
| $P(G,M)$, $G$ | The source's principal fibre bundle and its structural group, the "single geometrical structure" of the unified theory |
| $A$, $B$ | The two $2\times2$ halves of the source's array: $A$ the scalar half, $a_0$ times the identity, and $B$ the vector half, the three coefficients dotted into the three matrices |
| $\tilde Q = a_0 + \vec a\!\cdot\!\mathbf e$, $a_0 = p_0/m$, $a_k = i p_\eta/m$ | The biquaternion behind the source's array; Hermitian, of unit norm on the mass shell |
| $a_0^2+a_1^2+a_2^2+a_3^2 = 1$ | The source's Eq. (19); with imaginary $a_k$ it is $\langle\tilde Q\rangle_{\natural} = 1$ and the mass shell |
| $\langle\tilde Q\rangle_{\natural} = \tilde Q\tilde Q^{\natural}$ | The quaternionic norm, the sum of the squares of the four coefficients: equal to $1$ on the source's shell and to $-m^2$ for the corpus's material four-momentum in $\mathbb{B}_{\Phi}$ |
| $G = A\,I_4 + B\,J$, $J^2 = -I_4$ | The source's Eq. (4); $J$ is one basis operator, $L_{e_t}R_{\vec e_y}$ in its frame |
| $G^+$ | The source's conjugate transpose: the transpose with the quaternion sign. It is the ordinary transpose whenever the vector half is symmetric, and it equals the matrix adjoint on every array whose two halves are real |
| $G = a_0\,\mathrm{I} + \sum_\eta a_\eta\,iL_{\vec e_\eta}R_{\vec e_y}$ | The same group, basis operator for basis operator: the identity and the three spatial basis operators, with the coefficients the source's own |
| $G_2$ | The Majorana-conjugated group; real and symmetric |
| Eq. (28) | The source's four generators: the identity in the timelike slot, which is the slip, and the three spatial basis operators with a minus |
| $J$, $g$, the third tensor | The source's three invariant tensors: almost complex structure, "Lorentzian" metric, third fundamental tensor. The metric part does not reproduce: no symmetric form of inertia $(1,3)$ is invariant under the group, and the invariant forms have inertia $(4,0)$, $(0,4)$ or $(2,2)$. The inertia the source attaches to $g$, and the candidates searched before the computation, are in §The Legacy Form of the Source's Metric Claim |
| The Majorana transformation | The source's $*$-unitary involution, a sum of two basis operators: $\tfrac{1}{\sqrt2}\bigl(iL_{e_t}R_{\vec e_z}-L_{\vec e_y}R_{\vec e_x}\bigr)$ |
| $i\vec e_x$, $\vec e_y$, $i\vec e_z$ | The source's second triple, the split real form of the rotation triple; the middle element is the only one whose square is $-e_t$ |
| $(2{+}2)$ inertia | The source's second sign pattern: one sign flipped in the coefficient sum, $a_0^2+a_1^2-a_2^2+a_3^2 = 1$, i.e. $p_0^2-p_1^2+p_2^2-p_3^2 = m^{2}c^{2}$ |
| $M_2(\mathbb{H})$, $M_4(\mathbb{C})$ | The ambient operator algebra's real form, and the ambient itself: the corpus's reading of the source's quaternionic $2\times2$ algebra, whose complexification is the Dirac algebra |
| $L_a$, $R_b$, $L_aR_b$ | Left and right multiplication, $L_a(\tilde Q)=a\tilde Q$ and $R_b(\tilde Q)=\tilde Qb$, and the basis operator $L_aR_b(\tilde Q)=a\tilde Qb$ |
| Frame basis operators $C_\mu$ | $C_t=iR_{\vec e_x}$ and $C_\eta=iL_{\vec e_\eta}R_{\vec e_y}$, with $\{C_\mu,C_\nu\}=2\varepsilon_\mu\delta_{\mu\nu}$ and $\varepsilon=(1,-1,-1,-1)$ |
| $\mathsf{V}_{\mathrm{vol}}=-iL_{e_t}R_{\vec e_z}$ | The fifth basis operator, the chirality of the frame: its square is $+1$ and it anticommutes with the four frame basis operators |
| $\tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}$, $(\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}$ | The source's Eq. (31) read on the algebra: the two Dirac equations, i.e. the one-sided pair |
| The source's bar | The quaternion sign, $\tilde Q^{\natural}$; not the corpus's bar, which conjugates the coefficients |
Further Reading
- The AXF Matrices: Axis-Free, Integer Entries, for the reasons the AXF matrices replace the Pauli and gamma matrices.
- D. J. Cirilo-Lombardo, "Algebraic structures, physics and geometry from a Unified Field Theoretical framework", arXiv:1411.5493v3 [hep-th] (2015), for the source recorded here: the principal fibre bundle $P(G,M)$ and its structural group, the biquaternionic basis of the bundle, the Majorana and Dirac representations and their relation to the signs, the non-Hermitian CP operators, the split triple and the $(2{+}2)$ inertia, and the concluding geometry claims.
- J. Fredsted, "Spinor fields without Lorentz frames in curved spacetime using complexified quaternions," arXiv:0811.1357v4 [math-ph] (2009), for the constructed biquaternionic basis recorded in A Constructed Biquaternionic Basis, from an Independent Source: the material-sector basis $s_\mu \in \mathbb{B}_{\Phi}$ with the metric as the quaternionic form, the conjugation by a unit-norm element as the local Lorentz freedom, and the single algebra-valued connection whose imaginary-scalar and vector parts are a local $\mathrm{U}(1)$ and the Lorentz connection. Cited as an independent construction in the sector opposite to the source recorded here; the corpus does not adopt it, and it builds no principal bundle and no dynamics.
- Conventions in the Biquaternion Universe, for the algebra, the conjugations, the matrix representation $\mathsf{M}_2$ and the norm $\langle\cdot\rangle_{\natural}$ used throughout.
- Conventions of the Matrix Models, for the identification of the real form $M_2(\mathbb{H})$ and the ambient $M_4(\mathbb{C})$, §The Clifford Structure, which is the source's own justification for expecting a biquaternion.
- States, Observables and the Born Weight and Four-Vectors and the Interval, for the placement of the source's $\tilde Q$ in $\mathbb{B}_{\Psi}$ and of the corpus's four-momentum in $\mathbb{B}_{\Phi}$.
- The Neutrino and Majorana Fermions in Biquaternionic Form, for the corpus's sign-dependent classification of reality conditions and the Majorana condition on the doubled module.
- Curved Spacetime and the Biquaternion Framework, for the reasons the algebra cannot carry curvature, the field-of-frames route, and the conditions on the connection; and The Einstein Field Equations in the Biquaternion Framework: The Linearised Dictionary and the Non-Linear Gap for where a structure-group mechanism would be recorded.
- The Hopf Fibration and the Biquaternion Gauge Bundle, for the corpus's own principal bundle and the defining module as an associated bundle.
- Biquaternion Versors and the Orthogonal Group and The Grading and Fermion Parity in Biquaternionic Form, for the corpus's $\mathbb{Z}/2$ grading, the fermionic half of the source's interlude.
- A Brief History of Biquaternions in Physics and The Empirical Status of the Biquaternion Framework, for the corpus's standing on the algebra as a construction and on what could count as evidence that it is more than a choice.