The Path Integral and the Coupled Klein–Gordon Equation in Biquaternionic Polar Variables

Introduction

The integral is the companion's and this article reads it in the chart. The Path Integral and the Coupled Klein–Gordon Equation in Biquaternionic Form builds the functional integral over central fields with the covariant derivative in the exponent, splits that exponent into three terms, and identifies the operator the Gaussian inverts. This article substitutes the complete polar decomposition and reads where the connection lands.

The word on a central field is a scale and an angle. The boost and the rotor have nothing to carry and are $e_t$ and $\pm e_t$. So the chart has two functions per point, the measure has one Jacobian per point, and the connection has only two objects it can touch. The collapse is the polar equation article's and is not re-derived.

The finding is one sentence. The connection enters the exponent as a shift of one bracket, $(\partial_{r'_t}\theta_{\mathrm{pha}}-\frac{ϛ}{\hbar}a'_t)^{2}-\lvert\nabla\theta_{\mathrm{pha}}+\frac{ϛ}{\hbar}\vec A\rvert^{2}$, and the amplitude's two terms are untouched, so the two interaction terms of the element form are the expansion of that one shift and the chart shows their shapes.

Four consequences follow, and each is stated exactly or checked.

  • The element form's three-term split is the expansion of the chart's single shifted square, and the two interaction terms have the shapes $2c\varrho^{2}\partial\theta_{\mathrm{pha}}$ and $c^{2}\varrho^{2}$ and no other.
  • The measure is untouched, because its Jacobian is the scale and the scale is gauge invariant, so the chart's invariance needs no link variable and no argument.
  • The amplitude equation keeps the connection out of its quantum term, and the transport equation becomes the mechanical one; both are the two saddles, exactly as in the free article.
  • The chart's bracket is a squared modulus and therefore real for every connection, so its symbol is real; the equation's squared operator has an imaginary part wherever $\bar c_\mu\ne-c_\mu$. The chart carries the action's reading of the fork and cannot carry the equation's, and the reason is that the cross term the equation's reading needs has no chart counterpart.

What the article does not do. It does not produce a second integral, does not define the continuum measure, and does not decide the pole question of the element companion. It reads one integral in another coordinate system and states what the coordinate system can and cannot show.

Conventions

The algebra, the pairing and the word are the free polar article's: $\mathbb{B}=\mathbb{C}\otimes_{\mathbb{R}}\mathbb{H}$, $e_t=1$, $\vec e_\mu^{2}=-e_t$, $i$ central, $\mathrm{Sc}(\cdot^{*}\cdot)$, and $\tilde\Xi=\varrho e^{i\theta_{\mathrm{pha}}}\tilde U_{\Psi,\mathrm b}\tilde u_{\mathrm r}$ with the ranges of The Klein–Gordon Equation in Biquaternionic Polar Variables, §The Word.

The covariant derivative and its coefficient. With $c_\mu=\frac{iϙ}{\hbar}A_\mu$ the covariant derivative is $\partial_\mu+c_\mu$ on the coefficients, and for a central field

$$ \tilde\nabla_{\mathrm{gco},\Phi,\mu}\bigl(\varrho e^{i\theta_{\mathrm{pha}}}e_t\bigr) =e^{i\theta_{\mathrm{pha}}}\Bigl(\partial_\mu \varrho+i\,\varrho\,\kappa_{\mu}^{\mathrm{kin}}\Bigr)e_t, \qquad \kappa_{\mu}^{\mathrm{kin}}=\partial_\mu\theta_{\mathrm{pha}}+\frac{ϛ}{\hbar}A_\mu , $$

which is the coupled polar companion's factorisation. The derivative of the word is one real number and one shifted phase rate, and the shift is the whole of the connection's presence in the derivative.

The factorisation is checked and it converges at the second order. On a nine-site chain with a smooth scale, a smooth angle and one coefficient per slot, the two sides of $\mathrm{Sc}\bigl[(D_\mu\tilde\Xi)^{*}D_\mu\tilde\Xi\bigr]=(\partial_\mu \varrho)^{2}+\varrho^{2}(\kappa^{\mathrm{kin}}_\mu)^{2}$ differ by $2.2\times10^{-2}$, $5.6\times10^{-3}$, $1.4\times10^{-3}$ and $3.5\times10^{-4}$ relative, at steps $10^{-1}$, $5\times10^{-2}$, $2.5\times10^{-2}$ and $1.25\times10^{-2}$, falling by four at each halving. The identity is exact in the continuum and is a product rule away from it on a grid, which is the same status as the free article's $\lvert\partial\xi\rvert^{2}$ identity.

The exponent. The weight is $\exp(iS_{\mathrm{gco}}/\hbar)$ with the element companion's action, and in the chart

$$ \frac{i}{\hbar}S_{\mathrm{gco}}=\frac{i}{2}\int\mathrm{d}^{4}x\;\Bigl\{(\partial_{r'_t}\varrho)^{2}-\lvert\nabla \varrho\rvert^{2} +\varrho^{2}\Bigl[\bigl(\partial_{r'_t}\theta_{\mathrm{pha}}-\tfrac{ϛ}{\hbar}a'_t\bigr)^{2} -\bigl\lvert\nabla\theta_{\mathrm{pha}}+\tfrac{ϛ}{\hbar}\vec A\bigr\rvert^{2}\Bigr]-k_m^{2}\varrho^{2}\Bigr\}, $$

which is the free polar article's bracket with the two substitutions of the polar variational companion, §The Action in the Chart, and with the $\tfrac12$ of the free article's normalisation. The material convention is $A_t=i a'_t$ with $a'_t$ and $\vec A$ real, so the temporal shift carries a minus and the spatial one a plus; the two signs are the metric's and are the same two the polar equation companion records for the four-momentum.

The Exponent, and Where the Connection Sits

One object is shifted and three are not. The connection appears in the exponent only through $\kappa^{\mathrm{kin}}$. The amplitude's gradient $(\partial_{r'_t}\varrho)^{2}-\lvert\nabla \varrho\rvert^{2}$ is free of it, the density's weight $\varrho^{2}$ is free of it, and the mass term is free of it. The chart's exponent is the free exponent with one bracket replaced, and the replacement is a translation by a fixed function.

A translation of a bracket and not of a derivative. The shift is inside a square, so the exponent is not the free exponent plus a total derivative; the difference is the two interaction terms of the next section. The polar variational companion makes the same point from the variational side, and the phrase for it is that the connection's coefficient is central: a central element multiplies the word and cannot act inside a factor, so its only effect on a square is through the square's argument.

What the chart cannot show. No factor of the word other than the angle is moved, because the shift is added to $\partial\theta_{\mathrm{pha}}$ and to nothing else. In the element form the same fact is the statement that the connection's coefficients are central; in the chart it is the statement that the connection's only address is the phase's rate. The two statements are one statement, and the chart's version is the more specific because it names the coordinate.

The Two Interaction Terms and Their Shapes

The expansion. Squaring the shifted bracket gives three terms,

$$ \varrho^{2}\Bigl[\bigl(\partial_{r'_t}\theta_{\mathrm{pha}}\bigr)^{2}-\lvert\nabla\theta_{\mathrm{pha}}\rvert^{2}\Bigr] -\frac{2ϛ \varrho^{2}}{\hbar}\Bigl(\partial_{r'_t}\theta_{\mathrm{pha}}\,a'_t+\vec\nabla_\Phi\theta_{\mathrm{pha}}\cdot\vec A\Bigr) +\frac{ϛ^{2}\varrho^{2}}{\hbar^{2}}\Bigl(\bigl(a'_t\bigr)^{2}-\lvert\vec A\rvert^{2}\Bigr), $$

which is the free phase term, a term linear in the potential, and a term quadratic in it. The element companion's three-term split is the expansion of this one square, and the correspondence is term by term.

The shapes are fixed by the density. The linear term is the density times the phase rate times the potential, which is the current's slot: the polar companion's current is $\rho_{\mathrm{num}}\tilde P^{\mathrm{mech}}/m$ and its phase part is exactly $\rho_{\mathrm{num}}\kappa^{\mathrm{kin}}$. The quadratic term is the density times the potential squared, with the temporal component entering with $+$ and the three spatial ones with $-$ because the bracket is a difference of a temporal and a spatial square. The sign difference between the scalar and vector potential's squares is the metric's, and it is the chart's form of the element companion's remark that the two potentials are the two slots of one element.

The fourth term that is absent. The exponent contains no term linear in the potential and linear in the amplitude's gradient. The two interaction terms are $2c\varrho^{2}\partial\theta_{\mathrm{pha}}$ and $c^{2}\varrho^{2}$, and neither contains $\partial \varrho$. This absence is a structural fact about the chart and it decides the fork of the next but one section.

The quadratic term is not removable and it is what the invariance needs. The linear term alone changes by a total derivative under a gauge transformation, and the quadratic term is what supplies the remainder; the polar variational companion records the same cancellation as its third result. In the chart the gauge invariance is visible as an identity between two displayed terms, and this is sharper than the element form's statement, which has to combine the two terms' contributions in the exponent.

The Measure Is Untouched

The Jacobian is the scale and the scale is gauge invariant. The polar measure is $\prod_x \varrho\,\mathrm{d}\varrho\,\mathrm{d}\theta_{\mathrm{pha}}$, and a gauge transformation translates $\theta_{\mathrm{pha}}$ by $ϛ\Lambda/\hbar$ and leaves $\varrho$ alone. So the measure is invariant with no computation, whereas the element form has to note that the transformation is a rotation at each point and that a rotation has unit determinant.

The two statements are one and the chart's is the shorter. A rotation of the plane preserves the flat measure and the polar measure's radial factor; the chart's statement is that the radial factor does not move, and the element form's is that the angular factor does not. Neither statement needs the connection, which is why neither needs a regulator to be true.

The lattice's exponential link is implied and not postulated. The chart's invariance is a statement about the exact measure, so it holds for a finite transformation and at any step size, and the lattice formulation it corresponds to is the one whose link carries the phase as an exponential. The chart does not need to say this, because the chart does not discretise the angle into a link; the element companion's check that the linearised link fails and the Wilson link succeeds is the discrete shadow of the chart's exact statement.

The finite transformation and the chart's range. A finite translation can carry the angle out of $(-\pi/2,\pi/2]$, and the polar equation companions record that the chart is then not closed. In the integral the angle is integrated over a full turn, so the translation is a translation of a circle coordinate and the chart's half-turn range is a two-chart cover of it. The integral's version of the caveat is that a finite gauge transformation relabels the cover and moves no weight.

The Amplitude Equation Is Unchanged and the Transport Is Mechanical

The two saddles, in the order of the free article. Stationarity in the scale at fixed angle gives

$$ \kappa^{\mathrm{kin}}\!\cdot\kappa^{\mathrm{kin}}\equiv\bigl(\partial_{r'_t}\theta_{\mathrm{pha}}\bigr)^{2}-\lvert\nabla\theta_{\mathrm{pha}}\rvert^{2} =k_m^{2}-\frac{\Box_\Phi \varrho}{\varrho}, $$

which is the polar equation companion's amplitude equation with the mechanical bracket in place of the free one. Read through $S_{\tilde\Xi}=\hbar\theta_{\mathrm{pha}}$ it is $(\partial_{r'_t}S_{\tilde\Xi}-ϛa'_t)^{2}-\lvert\nabla S_{\tilde\Xi}+ϛ\vec A\rvert^{2}=m^{2}c^{2}-\hbar^{2}\Box_\Phi \varrho/\varrho$, the mass shell with the mechanical momentum and the quantum potential.

The quantum potential is untouched. It is $-\frac{\hbar^{2}}{2m}\Box_\Phi \varrho/\varrho$, a functional of the scale alone, and the connection does not appear in it because the connection does not appear in the amplitude's terms. This is the chart's sharpest statement about the coupling: the coupling changes the relation between the phase's rate and the momentum, and it changes nothing about the curvature of the density.

Stationarity in the angle gives the mechanical transport equation. Varying $\theta_{\mathrm{pha}}$ gives $\sum_\mu\partial_\mu\bigl(\varrho^{2}\kappa_{\mathrm{kin},\mu}\bigr)=0$, which is the divergence of the density times the mechanical four-velocity. The free article's saddle is the same equation with the free rate, so the coupling's whole effect on the constraint is to replace the phase's gradient by the mechanical one.

The check is the free check with a link phase. On the same one-dimensional chain as the free article's, with the same weights and the same endpoints, adding link phases $-0.61$ and $0.29$ leaves the angle's integral equal to its saddle times the Fresnel factor to $1.2\times10^{-13}$, and the saddle satisfies the discrete divergence condition to $4\times10^{-16}$ against a link current of $0.28$. The phase integral is its saddle times a Fresnel factor, and the saddle is the mechanical transport equation.

What the saddle is not. It is not the current's conservation in the sense of the element companion's Ward identity, which is a statement about every correlation function. The saddle is the statement about the dominant configuration, and the two agree on shell, as they must.

Which Reading the Chart Carries

The fork, in one sentence. The element companion distinguishes two operators: the action's paired operator $\sum_\mu(\partial_\mu-\bar c_\mu)(\partial_\mu+c_\mu)$ and the equation's squared operator $\sum_\mu(\partial_\mu+c_\mu)^{2}$, which coincide exactly when $\bar c_\mu=-c_\mu$ in every slot, hence for a purely vector connection, and not for a scalar potential.

The chart's bracket is the action's. It is a squared modulus, $\lvert(\partial_\mu+c_\mu)\Xi\rvert^{2}=(\partial_\mu \varrho)^{2}+\varrho^{2}(\kappa^{\mathrm{kin}}_\mu)^{2}$, so the exponent is real for every connection and the quadratic form is real. The chart writes the action's object and nothing else, and the reason is that a chart of a modulus can write a squared modulus and cannot write a square of an operator.

The equation's reading needs a term the chart does not have. The squared operator's real part carries a term proportional to $c_\mu\partial_\mu \varrho$, the amplitude–potential cross term, and §The Two Interaction Terms and Their Shapes shows that the chart's exponent has no such term: its two interaction terms are $2c\varrho^{2}\partial\theta_{\mathrm{pha}}$ and $c^{2}\varrho^{2}$. The absence is exact and it is not an approximation: the chart has two variables and one of them, the angle, carries the whole of the connection, so a term multiplying the potential by the other variable's gradient cannot appear.

The consequence is measured. On a periodic grid the symbol of the action's form is real for every coefficient, its imaginary part $1.6\times10^{-17}$; the symbol of the squared form has an imaginary part $2\Sigma c$ in the slot where $\bar c\ne-c$, measured $1.86$ against $2\times2.9973\times0.31=1.858$. A complex symbol is not a dispersion relation, so the squared reading's slot is the one that needs the total-derivative bookkeeping the element companion records as an open question.

What the chart therefore decides and does not decide. It decides which reading is the geometric one: the action's, because the chart can only write it. It does not decide which is correct physics; that is the element companion's open question, and the chart's contribution is to show that the two are not two coordinate systems of one object but two objects, of which only one is a squared modulus.

A Constant Background, in the Chart

The shift, and what it does to the bracket. For a constant potential the two shifts are constants, so the bracket is the free bracket with $\partial\theta_{\mathrm{pha}}$ translated by a constant and $\vec\nabla_\Phi\theta_{\mathrm{pha}}$ translated by a constant vector. A translation of a gradient is a change of the plane wave's wavevector and nothing else, which is the chart's form of the element companion's statement that a constant potential shifts the wavevector.

The scalar potential enters squared in the chart. Because the temporal slot's shift sits inside a square with a relative sign, a constant $a'_t$ contributes $+\frac{ϛ^{2}}{\hbar^{2}}a_t'^{2}\varrho^{2}$ to the exponent and no term linear in $\partial_{r'_t}\theta_{\mathrm{pha}}$. The chart's dispersion for a constant scalar potential has the potential squared and no linear shift, which is the element companion's paired shell. The linear shift belongs to the squared reading and has no chart expression, for the reason of the previous section.

The vector potential enters as a translation and the two are different objects. A constant $\vec A$ translates $\vec\nabla_\Phi\theta_{\mathrm{pha}}$, so it moves the wavevector and changes the bracket by the translation's square; a constant $a'_t$ does the same in the temporal slot. The chart shows the two as one operation, a translation, and the difference between them as the relative sign of the two squares. This is the chart's version of the element companion's contrast between the exact scalar shift and the vector potential's Landau problem, and the chart states the contrast as a sign.

What the chart does not show. It does not show the spectrum. A constant vector potential's Landau levels require the transverse structure of the wavefunction, and the chart's exponent is a function of two scalars; the levels are the polar equation companion's and are not re-derived. The chart locates the potential in one coordinate and leaves the spectrum to the equation.

What the Polar Reading Adds and Does Not Add

Adds.

  • The connection's address, named: the phase's rate and nothing else.
  • The element form's three-term split as the expansion of one shifted square, with the two interaction terms' shapes and the absence of an amplitude–potential cross term.
  • The measure's invariance as one sentence, with no link variable and no determinant.
  • The two saddles with the mechanical bracket, and the quantum potential's independence of the connection as a consequence of the amplitude's terms being untouched.
  • The fork's resolution in geometric terms: the chart can write the action's operator and cannot write the equation's, and the missing object is one cross term.
  • The constant background as a translation of a gradient, and the scalar and vector potentials as one operation with a relative sign.

Does not add. No new value, no new prediction, no continuum measure, no decision on the pole question, no spectrum, and no treatment of the interacting theory. The article is a reading, and its content is the location of the connection and the limit of what a chart of a modulus can express.

The honest ledger. The chart contributes one sentence about where the connection sits, one about what the measure does, one about the saddles, and one about the fork. It contributes nothing to the value of the integral, and it contributes a negative result of some weight: the equation's reading of the covariant square is not a reading this chart can express, which is a statement about the two readings and not about the chart.

Open Questions

  1. The cross term's absence and the corpus's equation. The chart shows that the action's reading has no amplitude–potential cross term and the equation's does. Whether the corpus's coupled equation should be stated with the paired operator, which the chart writes, or with the squared one, which it does not, is the element companion's open question and this article does not decide it. The chart's contribution is that the two readings are not two coordinate systems of one object.

  2. The winding sectors with a connection. The angle is compact and the mechanical bracket contains the potential, so a loop's winding and the enclosed flux enter the same term. Whether the polar integral in a fixed winding sector with a connection is a distinct object, and whether the theta-angle structure of a compact scalar follows, is not treated.

  3. The constraint's determinant with a connection. The angle's integral produces the inverse square root of the determinant of $-\partial_\mu \varrho^{2}\partial_\mu$ with the connection inside. Whether that determinant's scale dependence cancels against the radial integral, as the free article's does, is not shown.

  4. The interaction. A central interaction $\lambda \varrho^{4}$ leaves the chart's structure and the connection's address unchanged. The article does not pursue it.

  5. The two-chart cover and finite transformations. The article states that a finite gauge transformation relabels the cover and moves no weight. Whether the cover has an observable consequence — a sign in a correlation function of the word — is the free polar article's question and remains open.

Summary

The coupled Klein–Gordon integral read in the chart is the free one with one bracket translated. The connection enters as $\kappa^{\mathrm{kin}}=\partial\theta_{\mathrm{pha}}+\frac{ϛ}{\hbar}A$, the amplitude's two terms and the mass term are untouched, and the exponent's two interaction terms are $2c\varrho^{2}\partial\theta_{\mathrm{pha}}$ and $c^{2}\varrho^{2}$. The element companion's three-term split is the expansion of this one shift.

The measure is invariant for one reason: its Jacobian is the scale and the scale is gauge invariant. The two saddles are the amplitude equation with the mechanical bracket and the mechanical transport equation, both exact, the second verified to $10^{-13}$ against its Fresnel factor and to $10^{-16}$ against the discrete divergence condition. The quantum potential is untouched because the amplitude's terms are.

The chart carries the action's reading of the covariant square and cannot carry the equation's, because the equation's reading needs a cross term between the potential and the amplitude's gradient and the chart's exponent has none. The measured symbol of the action's form is real for every connection; the squared form's is not, its imaginary part being $2\Sigma c$ in the slot where $\bar c\ne-c$. That is the chart's contribution to the fork: the two readings are two objects, and only one of them is a squared modulus.

Summary of Notation

Symbol Meaning
$\tilde\Xi=\varrho e^{i\theta_{\mathrm{pha}}}e_t$ The word on a central field
$c_\mu=\frac{iϙ}{\hbar}A_\mu$ The slot's coefficient; $c_t$ real, $c_\eta$ imaginary
$\kappa^{\mathrm{kin}}_\mu=\partial_\mu\theta_{\mathrm{pha}}+\frac{ϛ}{\hbar}A_\mu$ The mechanical phase rate; the connection's only address
$\tilde\nabla_{\mathrm{gco},\Phi,\mu}(\varrho e^{i\theta_{\mathrm{pha}}}e_t)=e^{i\theta_{\mathrm{pha}}}\bigl(\partial_\mu \varrho+i \varrho\kappa^{\mathrm{kin}}_\mu\bigr)e_t$ The exact factorisation of the covariant derivative
$\prod_x \varrho\,\mathrm{d}\varrho\,\mathrm{d}\theta_{\mathrm{pha}}$ The polar measure; unchanged by a gauge transformation
$2c\varrho^{2}\partial\theta_{\mathrm{pha}}$ The linear interaction term; the current's slot
$c^{2}\varrho^{2}$ The quadratic interaction term; no derivative
$\frac{ϛ^{2}\varrho^{2}}{\hbar^{2}}\bigl((a'_t)^{2}-\lvert\vec A\rvert^{2}\bigr)$ The quadratic term in the material slots; the relative sign is the metric's
$\kappa^{\mathrm{kin}}\!\cdot\kappa^{\mathrm{kin}}=k_m^{2}-\Box_\Phi \varrho/\varrho$ The amplitude equation; the scale's saddle
$\sum_\mu\partial_\mu\bigl(\varrho^{2}\kappa_{\mathrm{kin},\mu}\bigr)=0$ The mechanical transport equation; the angle's saddle
$-\frac{\hbar^{2}}{2m}\Box_\Phi \varrho/\varrho$ The quantum potential; independent of the connection
$\sum_\mu(\partial_\mu-\bar c_\mu)(\partial_\mu+c_\mu)$ The action's paired operator; the chart's object
$\sum_\mu(\partial_\mu+c_\mu)^{2}$ The equation's squared operator; needs a cross term the chart lacks
$2\Sigma c$ The squared operator's symbol's imaginary part where $\bar c\ne-c$
$k_m=mc/\hbar$ The mass parameter

Further Reading