Global information in the Green–Schwarz response
Green and Schwarz showed how suitable local interactions can cancel gauge and gravitational anomalies in ten-dimensional supergravity coupled to Yang–Mills for the appropriate gauge groups [1 ] . Global information requires a further description. In the anomaly-line framework, curvature measures the local anomaly, while a flat line can retain nontrivial holonomy [2 §2.2] . The scalar studied here compares a normalized Green–Schwarz section with a spin reference; a nontrivial comparison does not by itself signal a failure of anomaly cancellation.
Evaluating this normalized response on a closed background requires global data even when the background gauge connections on the ten-manifold are trivial. Tachikawa–Zhang identified a nontrivial third-root response in their comparison of heterotic Green–Schwarz terms [3 ] . Quaternionic sphere bundles provide a family of ten-manifolds on which the response can be computed and its dependence on the background made explicit. The calculation relates a local Green–Schwarz integral and a closed fermion eta phase to one integral relative characteristic number. Its value can change with the integral String datum while the underlying ten-manifold and the fibre period remain fixed. Related worldsheet developments include the tertiary invariant of two-dimensional minimally supersymmetric field theories [4 ] and the study of non-supersymmetric heterotic branes and their Green–Schwarz pairings [5 ] . Wada–Yonekura [6 ] , which appears simultaneously with this paper, evaluate discrete Green–Schwarz couplings by a complementary worldsheet/tertiary-invariant method, including an order-three response on Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) . The present calculation instead uses a spacetime sphere-bundle construction and a twelve-dimensional APS filling to evaluate the normalized response.
The mathematical derivation of the sphere-bundle formula and the examples below are due to Li–Sun, in their unpublished working manuscript Green–Schwarz scalars on quaternionic sphere bundles [7 ] . The present companion explains the background data, calculation and physical consequences of those results for readers familiar with anomalies but not with differential cohomology or the Crowley–Goette invariant. Its purpose is to make the local integral, the closed spectral contribution and their common relative pairing visible in one calculation, while retaining the different premises of the physical and generalized-cohomology comparisons.
For a closed ten-manifold M M M with differential String data and prescribed trivial gauge connections and frames, the observable is an exponentiated scalar O ( M ) = exp ( 2 π i ϕ ( M ) ) \mathcal{O}(M)=\exp(2\pi i\phi(M)) O ( M ) = exp ( 2 π i ϕ ( M )) , with additive value ϕ ( M ) ∈ R / Z \phi(M)\in\mathbb{R}/\mathbb{Z} ϕ ( M ) ∈ R / Z . Its normalization compares the Green–Schwarz section to a spin reference with trivial framed gauge bundles. Let N N N be an eleven-dimensional filling carrying the gauge and differential trivialization data, with three-form component H 3 H_3 H 3 , and let W r e f W_{\mathrm{ref}} W ref be the spin reference. The local eight-form Y 8 Y_8 Y 8 is fixed by the full real fermion anomaly polynomial. The additive phase combines the local integral with a closed eta term:
ϕ ( M ) = − a N − μ ∗ ( Z ) , a N = ∫ N H 3 Y 8 , Z = N ‾ ∪ M W r e f . (1) \phi(M)=-a_N-\mu_*(Z),\qquad a_N=\int_N H_3Y_8,\qquad Z=\overline{N}\cup_M W_{\mathrm{ref}}. \tag*{(1)} ϕ ( M ) = − a N − μ ∗ ( Z ) , a N = ∫ N H 3 Y 8 , Z = N ∪ M W ref . (1)
The orientation-reversed filling and reference form the closed spin/gauge double Z Z Z . The closed term μ ∗ \mu_* μ ∗ uses a real normalization of the Atiyah–Patodi–Singer eta invariant of the specified virtual real fermion coefficient, with ξ = ( η A P S + dim C ker ) / 2 \xi=(\eta_{\mathrm{APS}}+\dim_{\mathbb{C}}\ker)/2 ξ = ( η APS + dim C ker ) /2 and the factor − 1 / 2 -1/2 − 1/2 multiplying ξ \xi ξ , as well as the specified finite tangent-connection transgression. The kernel term, reference and the two minus signs are part of the observable. § 2 gives the complete definition in (5) .
The scalar describes the specified anomaly response. Its identification with the Spin ( 32 ) / Z 2 \operatorname{Spin}(32)/\mathbb{Z}_2 Spin ( 32 ) / Z 2 heterotic Green–Schwarz phase divided by the E 8 × E 8 E_8\times E_8 E 8 × E 8 phase uses the published fermion and Green–Schwarz prescription: the same chirality adjoint gauginos, identical gravitational and ghost sectors, the normalized section and reference pairing, and no additional independent invertible sector. Section 9 states this input and compares the actual closed real Dirac systems. Equality of local polynomials would leave precisely that global identification undetermined. The response studied here does not specify a complete partition function or a microscopic realization.
The geometry makes its global inputs concrete. Let B B B be a closed connected spin seven-manifold, with specified Spin structure, metric and a quaternionic line V V V with connection. Its unit sphere bundle M = S ( V ) M=S(V) M = S ( V ) has dimension ten. If U U U is the underlying complex rank-two bundle, write c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) and λ = p 1 / 2 \lambda=p_1/2 λ = p 1 /2 . The extra base datum is an integral differential trivialization h ˇ \check{h} h ˇ , with three-form component h h h , satisfying
d h ˇ = λ ˇ ( T B ) + c ˇ , d h = λ ( T B ) + c . (2) \mathrm{d}\check{h}=\check{\lambda}(TB)+\check{c},\qquad\mathrm{d}h=\lambda(TB)+c. \tag*{(2)} d h ˇ = λ ˇ ( T B ) + c ˇ , d h = λ ( T B ) + c . (2)
The second equation is the curvature equation; the first also retains the integral nullhomotopy and its specified morphisms. A primitive of the four-form alone would not retain the integer information used by the response. This datum induces the differential String datum s ˇ M \check{s}_M s ˇ M on S ( V ) S(V) S ( V ) . All gauge bundles used to construct the filling lift to the actual common cover Spin ( 16 ) × Spin ( 16 ) \operatorname{Spin}(16)\times\operatorname{Spin}(16) Spin ( 16 ) × Spin ( 16 ) .
The disk bundle N = D ( V ) N=D(V) N = D ( V ) provides the induced boundary data and the local integral. With the base-then-fibre orientation and the outward boundary orientation used throughout, that integral is a N = 1 3 ∫ B h c a_N=\frac{1}{3}\int_B hc a N = 3 1 ∫ B h c and is generally nonzero. Computing only the eta term would therefore miss the dependence on h ˇ \check{h} h ˇ . To evaluate the closed term in (1) , extend the actual pair ( B , V ) (B,V) ( B , V ) over a spin eight-manifold W W W with a quaternionic line V ~ \widetilde{V} V . Associating the doubled four-dimensional fibre over W W W produces an auxiliary twelve-dimensional spin/gauge filling. An ordinary Atiyah–Patodi–Singer calculation [8 ] , including real index parity and the finite connection correction, evaluates the closed eta phase. This auxiliary filling carries the Spin and gauge data required by the index theorem, but its anomaly four-cocycle is not trivialized. That distinction allows the scalar to be nontrivial.
The specified integral boundary nullhomotopy determines a relative class z z z , recording the bulk characteristic class and its chosen boundary trivialization. Its cohomology group and the map to absolute cohomology are
z ∈ H 4 ( W , B ; Z ) , j : H 4 ( W , B ; Z ) ⟶ H 4 ( W ; Z ) . z\in H^4(W,B;\mathbb{Z}),\qquad j:H^4(W,B;\mathbb{Z})\longrightarrow H^4(W;\mathbb{Z}). z ∈ H 4 ( W , B ; Z ) , j : H 4 ( W , B ; Z ) ⟶ H 4 ( W ; Z ) .
The map j j j forgets the boundary trivialization. Put c W = c 2 ( U ~ ) c_W=c_2(\widetilde{U}) c W = c 2 ( U ) , where U ~ \widetilde{U} U underlies V ~ \widetilde{V} V , and λ W = λ ( T W ) \lambda_W=\lambda(TW) λ W = λ ( T W ) . The bulk integral and retained local term then combine into
ϕ ( S ( V ) , s ˇ M ) = K 3 ( m o d Z ) , \phi(S(V),\check{s}_M)=\frac{K}{3}\pmod{\mathbb{Z}}, ϕ ( S ( V ) , s ˇ M ) = 3 K ( mod Z ) ,
K = ⟨ c W ⌣ z , [ W , B ] ⟩ ∈ Z , j z = λ W + c W . (3) K=\langle c_W\smile z,[W,B]\rangle\in\mathbb{Z},\qquad jz=\lambda_W+c_W. \tag*{(3)} K = ⟨ c W ⌣ z , [ W , B ]⟩ ∈ Z , j z = λ W + c W . (3)
Here ⌣ \smile ⌣ is the cohomological cup product and [ W , B ] [W,B] [ W , B ] is the relative fundamental class. In forms, the pairing is ∫ W c W ( λ W + c W ) − ∫ B h ∧ c \int_W c_W(\lambda_W+c_W)-\int_B h\wedge c ∫ W c W ( λ W + c W ) − ∫ B h ∧ c ; the boundary subtraction is what turns the APS expression into this integral relative product. Changing W W W with all boundary data fixed changes K K K by a multiple of 24, and K K K is even. Thus the normalized response is a third root of unity on this sphere-bundle domain. The relative integer is the output of the characteristic calculation; no charge interpretation is needed to reach this conclusion.
The relative pairing connects the response to an established secondary invariant of the base. If H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 H^{3}(B;\mathbb{Q})=H^{4}(B;\mathbb{Q})=0 H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 , Crowley–Goette’s invariant t B ( V ) t_B(V) t B ( V ) is defined and their extrinsic formula has the same relative pairing with a different normalization [9 §1, Proposition 1.3] :
t B ( V ) = K 24 , ϕ ( S ( V ) , s ~ M ) = 8 t B ( V ) in R / Z . (4) t_B(V)=\frac{K}{24},\qquad\phi(S(V),\tilde{s}_M)=8t_B(V)\ \text{in }\mathbb{R}/\mathbb{Z}. \tag*{(4)} t B ( V ) = 24 K , ϕ ( S ( V ) , s ~ M ) = 8 t B ( V ) in R / Z . (4)
These two rational-cohomology vanishings are the exact additional domain restriction; simple connectedness is not required. They make the rational relative lift unique, while the integral differential data remain present. The formula for ϕ \phi ϕ in (3) applies more generally and does not define an extension of t t t outside its domain.
For the original Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) background, an actual quaternionic-line filling yields K = − 2 K=-2 K = − 2 , hence ϕ = 1 / 3 \phi=1/3 ϕ = 1/3 and O = exp ( 2 π i / 3 ) \mathcal{O}=\exp(2\pi i/3) O = exp ( 2 π i /3 ) . Recovering the published value requires the connection and framing transport, in addition to this number. The general sphere construction has outward fibre period − 2 -2 − 2 . The right-framed period + 1 +1 + 1 of the original background is obtained by the special inversion and differential transport on Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) ; it is not a generic reidentification of the fibre datum.
The larger domain exposes information that neither the ten-manifold nor its fibre period specifies. At fixed metric and gauge connection, changing h ~ \tilde{h} h ~ by a degree-three differential character with integral characteristic class α ∈ H 3 ( B ; Z ) \alpha\in H^{3}(B;\mathbb{Z}) α ∈ H 3 ( B ; Z ) changes the scalar by − ⟨ α ⌣ c , [ B ] ⟩ / 3 -\langle\alpha\smile c,[B]\rangle/3 − ⟨ α ⌣ c , [ B ]⟩ /3 . Li–Sun’s construction [7 ] gives one fixed ten-manifold with choices s ~ n \tilde{s}_n s ~ n for which O ( S ( V ) , s ~ n ) = exp ( 4 π i n / 3 ) \mathcal{O}(S(V),\tilde{s}_n)=\exp(4\pi i n/3) O ( S ( V ) , s ~ n ) = exp ( 4 π in /3 ) , giving all three values for n = 0 , 1 , 2 n=0,1,2 n = 0 , 1 , 2 . These are distinct integral String data, rather than three vacua or a phase transition. The example lies outside the Crowley–Goette domain. It makes the role of the larger-domain formula tangible: the response can distinguish global background choices that share the same underlying manifold and fibre period.
The same relative formula also has a six-dimensional reading. The virtual complex Weyl coefficient 4 U − 8 4U-8 4 U − 8 , where the integers denote bundle multiplicities, has anomaly density c ( λ + c ) / 3 c(\lambda+c)/3 c ( λ + c ) /3 . The gauge-dependent part of Lee–Tachikawa’s primitive half-doublet density is c ( λ + c ) / 24 c(\lambda+c)/24 c ( λ + c ) /24 , and their response uses backgrounds with d H = c dH=c d H = c [10 §3.3] . The factor of eight relates these normalizations; the different equations d h = λ + c dh=\lambda+c d h = λ + c and d H = c dH=c d H = c retain different background data. The resulting divisibility condition and the integral lattice factorization are explained within the established Green–Schwarz quantization framework of Monnier–Moore and Monnier–Moore–Park [11 , 12 ] , with no claim to construct a complete six-dimensional theory.
We first specify the normalized response and its background data in § 2 , then follow the disk construction and the direct APS calculation to the relative integer. The comparison, calibration and change law explain what that integer measures. Later sections describe the reduced six-dimensional response, the published heterotic prescription and the separately conditional comparison with integral topological modular forms (TMF) and vertex operator superalgebras. The latter requires actual compatible integral lifts and real equivariant characters, and uses the selected filling with zero local counterterm. These additional inputs are stated in § 10 ; the response calculation does not construct them.
Neither that conditional application nor the heterotic identification is needed to prove the sphere-bundle evaluation.
The normalized scalar and its background data
The Green–Schwarz response studied here is a phase assigned to a ten-dimensional spin manifold with a specified differential String datum. This datum includes an integral trivialization of the spin characteristic class λ = p 1 / 2 \lambda= p_1/2 λ = p 1 /2 , together with its refinement at the chosen connection. The response combines a local eleven-dimensional integral with a closed real fermion anomaly. The two contributions belong to the same normalization: specifying the anomaly polynomial fixes the local density, but does not by itself specify the real anomaly theory, its boundary line, or the reference against which that line is evaluated.
The connections and String trivialization are background inputs. The eleven- and twelve-dimensional manifolds used below are auxiliary geometric fillings for evaluating their response. Their introduction does not supply additional spacetime degrees of freedom. The issue is which global phase remains after a local anomaly density has been trivialized, so the integral and spectral definitions of that phase must be specified together.
For fermions with an anomaly, the partition function is naturally a vector in a one-dimensional anomaly line over the background. A trivialization of the anomaly supplies a unit vector in that line; its Hermitian pairing with the fermion vector is a number. This is the framework used by Tachikawa–Zhang [3 §4.2, equations (67)–(73)] . The scalar below extracts the Green–Schwarz factor by pairing with a spin reference having trivial gauge fields. In the general invertible-field-theory formulation, an anomaly assigns a graded line to a ten-dimensional background and a phase to a closed eleven-manifold [2 §2.3] . The line used here is the ordered graded tensor product of inverse-Pfaffian lines for the real fermion coefficient specified below. Invertibility refers to the anomaly sector; the fermion path integral can itself vanish in the presence of zero modes [2 §2.1] . The scalar is an anomaly response with specified background data, rather than a calculation of the magnitude or other sectors of a full partition function. The real coefficient, spectral convention and reference pairing fixed below determine its mathematical normalization [7 §2] . The physical identification with the published heterotic prescription is addressed separately in § 9 .
Let N N N be an eleven-dimensional filling of M M M carrying the integral and differential anomaly-trivializing data, and let W r e f W_{\mathrm{ref}} W ref be a spin filling of the same boundary tangent geometry with trivial framed gauges. For this scalar the boundary gauge connections on M M M are trivial in prescribed frames. Both fillings have product collars, and the boundary identifications include their Spin lifts. Their glued closed spin/gauge manifold is Z = N ‾ ∪ M W r e f Z = \overline{N} \cup_M W_{\mathrm{ref}} Z = N ∪ M W ref , where the bar reverses orientation. With the conventions developed below, define
O ( M ) = exp ( 2 π i ϕ ( M ) ) , ϕ ( M ) = − a N − μ ∗ ( Z ) , a N = ∫ N H 3 Y 8 , (5) \mathcal{O}(M) = \exp(2\pi i\phi(M)), \qquad\phi(M) = -a_N - \mu_*(Z), \qquad a_N = \int_N H_3Y_8, \tag*{(5)} O ( M ) = exp ( 2 π i ϕ ( M )) , ϕ ( M ) = − a N − μ ∗ ( Z ) , a N = ∫ N H 3 Y 8 , (5)
with ϕ ∈ R / Z \phi\in\mathbb{R}/\mathbb{Z} ϕ ∈ R / Z . Here H 3 H_3 H 3 is the three-form part of the trivialization, Y 8 Y_8 Y 8 is fixed by the full fermion coefficient, and μ ∗ \mu_* μ ∗ is the real reduced eta value at the specified connection. The spin reference need not carry an extension of the String datum. Keeping these two sorts of filling distinct will permit the nonzero response computed in § 4 .
The common cover and the full real coefficient
Put λ = p 1 / 2 \lambda= p_{1}/2 λ = p 1 /2 , and write T T T for the tangent bundle. The gauge backgrounds are actual principal Spin ( 16 ) × Spin ( 16 ) \operatorname{Spin}(16) \times\operatorname{Spin}(16) Spin ( 16 ) × Spin ( 16 ) bundles, with associated rank-sixteen real vector bundles E , E ′ E,E' E , E ′ . This is the common cover on which the comparison is made. A background for a quotient gauge group that does not lift to this cover is outside the present domain. In particular, the cover matters when using the bordism calculation of Basile–Debray–Delgado–Montero [13 §3.3.3, Theorem 3.78] .
The virtual real coefficient and its characteristic forms are
Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , (6) \Delta_{\mathbb{R}} = E \otimes_{\mathbb{R}} E' - S_{\mathbb{R}}^{+}(E) - S_{\mathbb{R}}^{+}(E'), \tag*{(6)} Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , (6)
n = λ ( E ) , n ′ = λ ( E ′ ) , n = \lambda(E), \qquad n' = \lambda(E'), n = λ ( E ) , n ′ = λ ( E ′ ) ,
X 4 = λ ( T ) − n − n ′ , Y 8 = n 2 + n n ′ + n ′ 2 − p 2 ( E ) − p 2 ( E ′ ) 6 . (7) X_{4} = \lambda(T) - n - n', \qquad Y_{8} = \frac{n^{2} + nn' + n'^{2} - p_{2}(E) - p_{2}(E')}{6}. \tag*{(7)} X 4 = λ ( T ) − n − n ′ , Y 8 = 6 n 2 + n n ′ + n ′ 2 − p 2 ( E ) − p 2 ( E ′ ) . (7)
Here S R + S_{\mathbb{R}}^{+} S R + denotes the real positive half-spin representation, and Δ C \Delta_{\mathbb{C}} Δ C is the complexification of the virtual real coefficient. The two subtracted half-spin bundles each have real rank 128; subtraction denotes their inverse anomaly factors. Characteristic classes below also denote their Chern–Weil forms when connections are fixed. The real structure fixes the half factor in the anomaly polynomial. With A ^ ( T ) \widehat{A}(T) A ( T ) the Dirac index characteristic form and [ ⋅ ] 12 [\cdot]_{12} [ ⋅ ] 12 the form-degree twelve component, it reads
I 12 = − 1 2 [ A ^ ( T ) ch ( Δ C ) ] 12 = X 4 Y 8 . (8) I_{12} = -\frac{1}{2}\left[\widehat{A}(T)\operatorname{ch}(\Delta_{\mathbb{C}})\right]_{12} = X_{4}Y_{8}. \tag*{(8)} I 12 = − 2 1 [ A ( T ) ch ( Δ C ) ] 12 = X 4 Y 8 . (8)
We normalize the curvature F F F by absorbing 1 / ( 2 π ) 1/(2\pi) 1/ ( 2 π ) into it, so that ch = tr e i F \operatorname{ch} = \operatorname{tr} e^{iF} ch = tr e i F and c 2 = tr F 2 / 2 c_{2} = \operatorname{tr} F^{2}/2 c 2 = tr F 2 /2 for the symplectic complex representation [3 §2, §4.3] .
The factorization can be read directly from the character expansion before specializing the gauge bundles. Set S = n 2 + n n ′ + n ′ 2 − p 2 ( E ) − p 2 ( E ′ ) S = n^{2} + nn' + n'^{2} - p_{2}(E) - p_{2}(E') S = n 2 + n n ′ + n ′ 2 − p 2 ( E ) − p 2 ( E ′ ) . The vector and half-spin characters give the following components, with subscripts denoting form degree:
ch 0 ( Δ C ) = ch 4 ( Δ C ) = 0 , ch 8 ( Δ C ) = 4 S , ch 12 ( Δ C ) = n + n ′ 3 S . \operatorname{ch}_{0}(\Delta_{\mathbb{C}}) = \operatorname{ch}_{4}(\Delta_{\mathbb{C}}) = 0, \qquad\operatorname{ch}_{8}(\Delta_{\mathbb{C}}) = 4S, \qquad\operatorname{ch}_{12}(\Delta_{\mathbb{C}}) = \frac{n+n'}{3}S. ch 0 ( Δ C ) = ch 4 ( Δ C ) = 0 , ch 8 ( Δ C ) = 4 S , ch 12 ( Δ C ) = 3 n + n ′ S .
Since A ^ 4 = − λ ( T ) / 12 \widehat{A}_{4} = -\lambda(T)/12 A 4 = − λ ( T ) /12 , the degree-twelve product is ( n + n ′ − λ ( T ) ) S / 3 (n+n'-\lambda(T))S/3 ( n + n ′ − λ ( T )) S /3 . Multiplication by − 1 / 2 -1/2 − 1/2 produces (8) , including both gauge p 2 p_{2} p 2 terms. Thus the coefficient is a statement about the representations and their real normalization, not merely their net rank. The character expansion and the real parity argument are completed in Appendix § 14 .
Integral differential trivialization
The topological structure records an integral trivialization of the class represented by X 4 X_{4} X 4 . Here B G BG B G denotes the classifying space of a group G G G , and K ( Z , 4 ) K(\mathbb{Z},4) K ( Z , 4 ) is the space classifying degree-four integral cohomology. The characteristic-class map sends the tangent and gauge bundles to λ T − λ E − λ E ′ \lambda_T-\lambda_E-\lambda_{E'} λ T − λ E − λ E ′ :
B H = hofib ( B Spin × B Spin ( 16 ) × B Spin ( 16 ) → λ T − λ E − λ E ′ K ( Z , 4 ) ) . (9) B\mathcal{H}=\operatorname{hofib}\left(B\operatorname{Spin}\times B\operatorname{Spin}(16)\times B\operatorname{Spin}(16)\xrightarrow{\lambda_T-\lambda_E-\lambda_{E'}}K(\mathbb{Z},4)\right). \tag*{(9)} B H = hofib ( B Spin × B Spin ( 16 ) × B Spin ( 16 ) λ T − λ E − λ E ′ K ( Z , 4 ) ) . (9)
The homotopy fibre records the Spin and gauge bundles together with an integral nullhomotopy of their degree-four class. A differential H \mathcal{H} H -structure refines this datum using the connections and a three-form satisfying d H 3 = X 4 dH_3=X_4 d H 3 = X 4 . The curvature equation is a necessary part of the background, but loses the integral information if treated as its entire definition.
This is the global counterpart of the familiar Green–Schwarz Bianchi identity. In a geometric description of the heterotic B B B -field, a specified homotopy joins the gravitational and gauge characteristic cocycles, and its differential refinement supplies the three-form field strength [14, §2.2.1, (2.10)–(2.11); §3.2, (3.23)]. Here the characteristic class and common cover are fixed by (9) ; the following cocycle equations make their trivializing data explicit.
For example, represent the degree-four differential cocycle by X ˇ = ( c X , h X , X 4 ) \check{X}=(c_X,h_X,X_4) X ˇ = ( c X , h X , X 4 ) , where c X c_X c X is an integral four-cochain, h X h_X h X a real three-cochain, and I \mathcal{I} I integrates forms over singular chains. Write δ \delta δ for the singular-cochain coboundary and d d d for the exterior derivative on forms. A geometric trivialization is a triple ( b , k , H 3 ) (b,k,H_3) ( b , k , H 3 ) obeying
δ b = c X , I ( H 3 ) − b − δ k = h X , d H 3 = X 4 , (10) \delta b=c_X,\qquad\mathcal{I}(H_3)-b-\delta k=h_X,\qquad dH_3=X_4, \tag*{(10)} δ b = c X , I ( H 3 ) − b − δ k = h X , d H 3 = X 4 , (10)
with b b b integral of degree three and k k k real of degree two. These are the geometric trivialization equations of Redden [15 Propositions 3.11–3.12] . The middle relation ties the form to the integral nullhomotopy, with the integral cochain b b b included in the real cochains. Such a trivialization requires the integral characteristic class to vanish, although the curvature form X 4 X_4 X 4 need not be zero. It therefore does not identify a nonflat differential cocycle with the zero differential character. A real primitive of X 4 X_4 X 4 supplies only the last equation and cannot replace the full triple.
The isomorphism classes of geometric trivializations form a torsor for degree-three differential cohomology H ^ 3 ( − ; Z ) \widehat{H}^{3}(-;\mathbb{Z}) H 3 ( − ; Z ) ; the underlying integral trivialization classes form a torsor for H 3 ( − ; Z ) H^3(-;\mathbb{Z}) H 3 ( − ; Z ) [15 Theorem 3.15] . Thus two choices have a difference in the indicated group, but no preferred zero choice. Specified morphisms between differential cocycles also induce morphisms of their trivializations, and that induced identification depends on the chosen morphism [15 Proposition 3.17] . This is why the tautological boundary frame and the characteristic morphisms used in the disk construction must be retained. They identify the background on which the scalar is evaluated, rather than adding an optional description of the same curvature form.
On a boundary with trivial framed gauges the structure reduces to a differential String trivialization of λ ^ ( T M ) \widehat{\lambda}(TM) λ ( T M ) . Transport along an oriented differential String isomorphism preserves the scalar. An independent integral shift of the trivialization is a different input unless an actual isomorphism identifies it with the original one. § 7 will compute the resulting dependence on this global datum.
Real eta values and finite connection transport
On a closed spin/gauge eleven-manifold the spectral convention uses the Atiyah–Patodi–Singer (APS) eta invariant, which omits zero eigenvalues from the spectral asymmetry. Let D D D be the ordinary self-adjoint Levi–Civita Dirac operator, with its twisting coefficient indicated by the subscript:
ξ ( D ) = η A P S ( D ) + dim C ker D 2 , μ I ( Z ) = − 1 2 ξ ( D Z , Δ C ) ( m o d Z ) . (11) \xi(D)=\frac{\eta_{\mathrm{APS}}(D)+\dim_{\mathbb{C}}\ker D}{2},\qquad\mu_I(Z)=-\frac{1}{2}\xi(D_{Z,\Delta_{\mathbb{C}}})\pmod{\mathbb{Z}}. \tag*{(11)} ξ ( D ) = 2 η APS ( D ) + dim C ker D , μ I ( Z ) = − 2 1 ξ ( D Z , Δ C ) ( mod Z ) . (11)
For the virtual coefficient, the real-valued invariants of its summands are combined before reduction modulo Z \mathbb{Z} Z . Halving an arbitrary element of R / Z \mathbb{R}/\mathbb{Z} R / Z would leave an unresolved sign in its exponential. The half in (11) instead comes from the real Pfaffian theory. In dimension 11 ≡ 3 ( m o d 8 ) 11\equiv3\pmod8 11 ≡ 3 ( mod 8 ) , the Dirac eigenspaces for a real twist are quaternionic; this is the spectral structure behind the half-eta function and boundary Pfaffian line of Freed–Moore [16 Proposition 1.28, §§3.1–3.2] .
The kernel term is part of this normalization even when a chosen metric has zero modes. If h D = dim C ker D h_D=\dim_{\mathbb{C}}\ker D h D = dim C ker D , orientation reversal gives ξ ( − D ) = − ξ ( D ) + h D \xi(-D)=-\xi(D)+h_D ξ ( − D ) = − ξ ( D ) + h D . Quaternionic symmetry makes h D h_D h D even, so multiplication by − 1 / 2 -1/2 − 1/2 yields μ I ( − Z ) = − μ I ( Z ) \mu_I(-Z)=-\mu_I(Z) μ I ( − Z ) = − μ I ( Z ) in R / Z \mathbb{R}/\mathbb{Z} R / Z . Omitting the kernel term would discard the reduced invariant precisely at the point where its spectral definition requires it.
The same real structure fixes the period in the filling calculation. For a twelve-dimensional spin manifold Q Q Q with real twists and product boundary data, APS gives
ξ ( D ∂ Q , Δ C ) = ∫ Q A ^ ( T Q ) ch ( Δ C ) − Ind A P S D Δ C + . \xi(D_{\partial Q,\Delta_{\mathbb{C}}})=\int_Q\widehat{A}(TQ)\operatorname{ch}(\Delta_{\mathbb{C}})-\operatorname{Ind}_{\mathrm{APS}}D_{\Delta_{\mathbb{C}}}^{+}. ξ ( D ∂ Q , Δ C ) = ∫ Q A ( T Q ) ch ( Δ C ) − Ind APS D Δ C + .
In dimension 12 ≡ 4 ( m o d 8 ) 12\equiv4\pmod8 12 ≡ 4 ( mod 8 ) , the quaternionic structures commute with the Dirac operator and preserve the APS spectral projections. Both chiral APS kernels consequently have even complex dimension, and hence the virtual index is even. The half-index therefore disappears modulo Z \mathbb{Z} Z , leaving
μ I ( ∂ Q ) = ∫ Q I 12 ( m o d Z ) . (12) \mu_I(\partial Q)=\int_Q I_{12}\pmod{\mathbb{Z}}. \tag*{(12)} μ I ( ∂ Q ) = ∫ Q I 12 ( mod Z ) . (12)
The twisted APS index theorem supplies the underlying formula [8 Theorem (4.2), formula (4.3)] . Witten–Yonekura describe the real antilinear symmetry and its role in the half-index normalization [17 equations (3.8)–(3.9)] . This statement requires a spin/gauge filling, with no requirement that the geometric trivialization extend over Q Q Q .
The disk calculation uses a specified stable metric Spin connection A T A_T A T , which can differ from the Levi–Civita connection. We adopt the finite transgression convention
d T Q ( A 0 , A 1 ) = Q ( A 1 ) − Q ( A 0 ) , μ ∗ ( Z ; A T ) = μ I ( Z ; A L C ) + ∫ Z T I 12 ( A L C , A T ) ( m o d Z ) . (13) \mathrm{d}T_Q(A_0,A_1)=Q(A_1)-Q(A_0),\qquad\mu_*(Z;A_T)=\mu_I(Z;A_{\mathrm{LC}})+\int_Z T_{I_{12}}(A_{\mathrm{LC}},A_T)\pmod{\mathbb{Z}}. \tag*{(13)} d T Q ( A 0 , A 1 ) = Q ( A 1 ) − Q ( A 0 ) , μ ∗ ( Z ; A T ) = μ I ( Z ; A LC ) + ∫ Z T I 12 ( A LC , A T ) ( mod Z ) . (13)
with fixed gauge connections. In the first equality Q Q Q denotes a characteristic polynomial. The second defines the differential-index value at A T A_T A T ; it does not identify that value with the raw spectrum of a Dirac operator with torsion. Finite changes of connection thus carry a specified correction. Using the stable value in the local term but the uncorrected Levi–Civita value in the closed term would evaluate two different geometric prescriptions. The APS period likewise holds with μ ∗ \mu_\ast μ ∗ and the characteristic forms at that stable connection. Below the closed anomaly values use μ ∗ \mu_\ast μ ∗ , including this correction whenever it is required.
Real bordism maps and the normalized section
Let L ( M ) L(M) L ( M ) denote the ordered graded tensor product of the inverse Pfaffian lines for the real coefficient (6) . The tensor factors are those of the honest real summands, with duals for the subtracted summands. Appendix § 14 .§ 5 constructs their unit eta states using Freed–Moore’s real boundary conditions and Witten–Yonekura’s real-state gluing [16 §3.2] [17 §§2.5, 3.1] . It retains the spectral-cut phases and norms, zero modes, and actual Spin and gauge seam maps. For disconnected boundaries, orientation duality reverses the order of the dual tensor factors. If the original order is retained instead, the corresponding sign is ( − 1 ) ∑ r < s p r p s (-1)^{\sum_{r<s}p_rp_s} ( − 1 ) ∑ r < s p r p s , where p j p_j p j is the real Pfaffian parity of the j j j th component. Even total parity does not remove these signs.
Write τ Δ , ∗ ( Y ) ∈ L ( ∂ Y ) \tau_{\Delta,\ast}(Y)\in L(\partial Y) τ Δ , ∗ ( Y ) ∈ L ( ∂ Y ) for the resulting unit state at the specified stable tangent connection. Its transport is (B.19): the ordinary Levi–Civita state is multiplied by exp [ − 2 π i ∫ Y T I 12 ( A L C , A T ) ] \exp[-2\pi i\int_Y T_{I_{12}}(A_{\mathrm{LC}},A_T)] exp [ − 2 π i ∫ Y T I 12 ( A LC , A T )] , with the boundary lines identified along the prescribed connection homotopy. On a closed manifold we denote this value by
Z − ( Z ) = τ Δ , ∗ ( Z ) = exp ( − 2 π i μ ∗ ( Z ) ) . Z_-(Z)=\tau_{\Delta,\ast}(Z)=\exp(-2\pi i\mu_\ast(Z)). Z − ( Z ) = τ Δ , ∗ ( Z ) = exp ( − 2 π i μ ∗ ( Z )) .
The minus sign in the finite factor is required by (13) ; the state and the local integral use the same connection prescription.
For a bordism A : M 0 → M 1 A:M_0\to M_1 A : M 0 → M 1 , the ordered orientation-dual identifications place its state in L ( M 1 ) ⊗ ^ L ( M 0 ) ∗ L(M_1)\mathbin{\widehat{\otimes}}L(M_0)^* L ( M 1 ) ⊗ L ( M 0 ) ∗ , where ⊗ ^ \widehat{\otimes} ⊗ denotes the graded tensor product. Evaluation against the second factor defines a linear map
Z − ( A ) : L ( M 0 ) ⟶ L ( M 1 ) , Z − ( A ) x = ( id ⊗ ^ ev x ) τ Δ , ∗ ( A ) . (14) Z_-(A):L(M_0)\longrightarrow L(M_1),\qquad Z_-(A)x=(\operatorname{id}\mathbin{\widehat{\otimes}}\operatorname{ev}_x)\tau_{\Delta,\ast}(A). \tag*{(14)} Z − ( A ) : L ( M 0 ) ⟶ L ( M 1 ) , Z − ( A ) x = ( id ⊗ ev x ) τ Δ , ∗ ( A ) . (14)
Here ev x \operatorname{ev}_x ev x evaluates a covector on x x x with the prescribed graded order. This defines the map before any composition law is used. For a filling N : ∅ → M N:\varnothing\to M N : ∅ → M , the empty line is C \mathbb{C} C with unit 1 1 1 , and Z − ( N ) ( 1 ) = τ Δ , ∗ ( N ) Z_-(N)(1)=\tau_{\Delta,\ast}(N) Z − ( N ) ( 1 ) = τ Δ , ∗ ( N ) .
The composition needed on the component of H \mathcal{H} H -boundaries follows from the actual closed-gluing pairing. Take differential H \mathcal{H} H -fillings N i : ∅ → M i N_i:\varnothing\to M_i N i : ∅ → M i and a differential H \mathcal{H} H -bordism A : M 0 → M 1 A:M_0\to M_1 A : M 0 → M 1 . Pair both Z − ( A ) τ Δ , ∗ ( N 0 ) Z_-(A)\tau_{\Delta,\ast}(N_0) Z − ( A ) τ Δ , ∗ ( N 0 ) and τ Δ , ∗ ( A ∘ N 0 ) \tau_{\Delta,\ast}(A\circ N_0) τ Δ , ∗ ( A ∘ N 0 ) with the same unit cap τ Δ , ∗ ( N 1 ) \tau_{\Delta,\ast}(N_1) τ Δ , ∗ ( N 1 ) . The real contraction of Appendix § 14 .§ 5 , whose reference case is (B.20), gives
⟨ τ Δ , ∗ ( N 1 ) , Z − ( A ) τ Δ , ∗ ( N 0 ) ⟩ = Z − ( N ‾ 1 ∪ M 1 A ∪ M 0 N 0 ) = ⟨ τ Δ , ∗ ( N 1 ) , τ Δ , ∗ ( A ∘ N 0 ) ⟩ . (15) \begin{aligned}
\left\langle\tau_{\Delta,\ast}(N_1),Z_-(A)\tau_{\Delta,\ast}(N_0)\right\rangle
=Z_-\left(\overline{N}_1\cup_{M_1}A\cup_{M_0}N_0\right) \\
&=\left\langle\tau_{\Delta,\ast}(N_1),\tau_{\Delta,\ast}(A\circ N_0)\right\rangle. \tag*{(15)}
\end{aligned} ⟨ τ Δ , ∗ ( N 1 ) , Z − ( A ) τ Δ , ∗ ( N 0 ) ⟩ = Z − ( N 1 ∪ M 1 A ∪ M 0 N 0 ) = ⟨ τ Δ , ∗ ( N 1 ) , τ Δ , ∗ ( A ∘ N 0 ) ⟩ . (15)
For the first closed gluing, cap A A A by N ‾ 0 ∪ N 1 \overline{N}_0\cup N_1 N 0 ∪ N 1 , using reversed dual tensor order on N ‾ 0 \overline{N}_0 N 0 . For the second, cap A ∘ N 0 A\circ N_0 A ∘ N 0 by N 1 N_1 N 1 . Both give the displayed closed manifold with the same component order and actual seam identifications. The first slot of the Hermitian pairing is conjugate-linear, which reverses N 1 N_1 N 1 . Since the cap is nonzero and L ( M 1 ) L(M_1) L ( M 1 ) is one-dimensional, equal pairings imply equal vectors. The finite counterterm is additive across matched product collars, so the equality holds for the transported states as written, not only for their Levi–Civita values. This proves the required composition on these states without assuming an equivalence of complete anomaly functors.
For a differential H \mathcal{H} H -filling N N N define
v − ( M ) = exp ( 2 π i a N ) Z − ( N ) ( 1 ) . (16) v_-(M)=\exp(2\pi i a_N)Z_-(N)(1). \tag*{(16)} v − ( M ) = exp ( 2 π i a N ) Z − ( N ) ( 1 ) . (16)
This is a unit vector in L ( M ) L(M) L ( M ) . Its filling independence follows from the closed cancellation identity and the real contraction.
The cancellation of the local and spectral contributions is first a statement on closed manifolds. On a closed differential H \mathcal{H} H eleven-manifold C C C , the common-cover bordism result Ω 11 H = 0 \Omega^{\mathcal{H}}_{11}=0 Ω 11 H = 0 [13 Theorem 3.78] extends the integral structure over a twelve-manifold Q H Q_{\mathcal{H}} Q H . The metric and gauge connections extend with product collars. The forgetful map H ^ 3 ( Q H ; Z ) → H 3 ( Q H ; Z ) \widehat{\mathcal{H}}^3(Q_{\mathcal{H}};\mathbb{Z})\to H^3(Q_{\mathcal{H}};\mathbb{Z}) H 3 ( Q H ; Z ) → H 3 ( Q H ; Z ) is surjective, and the corresponding map of trivialization torsors is equivariant. It therefore gives a geometric refinement of the specified bulk integral trivialization [15 Theorem 3.15] . Its boundary refinement can differ from the prescribed one only by a differential character with zero characteristic class. Such a difference is represented by a global two-form on C C C . Extending that form into the collar and adjusting the bulk refinement makes the boundary agree up to the specified isomorphism; its cochain representative is retained along the collar [7 §2] . This extends the prescribed refinement after the integral nullhomotopy has been extended, without requiring surjectivity of H 3 ( Q H ; Z ) → H 3 ( C ; Z ) H^3(Q_{\mathcal{H}};\mathbb{Z})\to H^3(C;\mathbb{Z}) H 3 ( Q H ; Z ) → H 3 ( C ; Z ) . Thus (8) , (12) , and Stokes give
μ ∗ ( C ) = ∫ Q H X 4 Y 8 = ∫ C H 3 Y 8 ( m o d Z ) . (17) \mu_*(C)=\int_{Q_{\mathcal H}}X_4Y_8=\int_C H_3Y_8\pmod{\mathbb{Z}}. \tag*{(17)} μ ∗ ( C ) = ∫ Q H X 4 Y 8 = ∫ C H 3 Y 8 ( mod Z ) . (17)
Here Y 8 Y_8 Y 8 is closed. No assertion that Y 8 Y_8 Y 8 itself has integral periods is needed: the real APS formula controls the full exponential.
For two such fillings N 0 , N 1 N_0,N_1 N 0 , N 1 of the same specified boundary, let C = N 0 ‾ ∪ M N 1 C=\overline{N_0}\cup_M N_1 C = N 0 ∪ M N 1 , including their differential identifications. The actual real gluing and (17) give
⟨ v − ( M ; N 0 ) , v − ( M ; N 1 ) ⟩ = exp ( 2 π i ( a N 1 − a N 0 ) ) Z − ( C ) = exp ( 2 π i ∫ C H 3 Y 8 − 2 π i μ ∗ ( C ) ) = 1. \begin{aligned}
\left\langle v_-(M;N_0),v_-(M;N_1)\right\rangle=\exp\left(2\pi i(a_{N_1}-a_{N_0})\right)Z_-(C) \\
&=\exp\left(2\pi i\int_C H_3Y_8-2\pi i\mu_*(C)\right)=1.
\end{aligned} ⟨ v − ( M ; N 0 ) , v − ( M ; N 1 ) ⟩ = exp ( 2 π i ( a N 1 − a N 0 ) ) Z − ( C ) = exp ( 2 π i ∫ C H 3 Y 8 − 2 π i μ ∗ ( C ) ) = 1.
Both vectors have unit norm, so they coincide. This proves filling independence with the ordered real contraction, rather than by choosing a square root of a complex determinant identity.
Concatenate A A A with N 0 N_0 N 0 . The local integral is additive at the matched collars, and (15) identifies its eta state with Z − ( A ) Z − ( N 0 ) ( 1 ) Z_-(A)Z_-(N_0)(1) Z − ( A ) Z − ( N 0 ) ( 1 ) . Filling independence then gives
v − ( M 1 ) = exp ( 2 π i ∫ A H 3 Y 8 ) Z − ( A ) v − ( M 0 ) . (18) v_-(M_1)=\exp\left(2\pi i\int_A H_3Y_8\right)Z_-(A)v_-(M_0). \tag*{(18)} v − ( M 1 ) = exp ( 2 π i ∫ A H 3 Y 8 ) Z − ( A ) v − ( M 0 ) . (18)
Conversely, applying this law to a filling from the empty manifold fixes the section once v − ( ∅ ) = 1 v_-(\varnothing)=1 v − ( ∅ ) = 1 is fixed. Its uniqueness is on the stated component of boundaries of H \mathcal{H} H -fillings.
The framed spin reference and the scalar
Closed cancellation controls comparisons between H \mathcal{H} H -fillings; it does not force the scalar formed with a spin reference to be one. The reference has trivial framed gauges and matching boundary tangent geometry, but need not extend the differential H \mathcal{H} H datum. Thus N ‾ ∪ M W r e f \overline{N}\cup_M W_{\mathrm{ref}} N ∪ M W ref is a closed spin/gauge manifold to which the closed eta value applies; it need not be a closed differential H \mathcal{H} H manifold satisfying (17) .
For fixed framed trivial principal gauge connections, the actual parallel virtual coefficient cancels as 256 − 128 − 128 = 0 256-128-128=0 256 − 128 − 128 = 0 and Y 8 = 0 Y_8=0 Y 8 = 0 . Its real operators and kernels cancel with their inverse factors, fixing the canonical even reference line and unit. The ordered evaluation and the real Hermitian gluing (B.20) therefore give
⟨ v − ( M ) , Z − ( W r e f ) ( 1 ) ⟩ = exp ( − 2 π i a N ) ⟨ Z − ( N ) ( 1 ) , Z − ( W r e f ) ( 1 ) ⟩ = exp ( − 2 π i [ a N + μ ∗ ( N ‾ ∪ M W r e f ) ] ) . \left\langle v_-(M),Z_-(W_{\mathrm{ref}})(1)\right\rangle=\exp(-2\pi i a_N)\left\langle Z_-(N)(1),Z_-(W_{\mathrm{ref}})(1)\right\rangle
=\exp\left(-2\pi i\left[a_N+\mu_*\left(\overline{N}\cup_M W_{\mathrm{ref}}\right)\right]\right). ⟨ v − ( M ) , Z − ( W ref ) ( 1 ) ⟩ = exp ( − 2 π i a N ) ⟨ Z − ( N ) ( 1 ) , Z − ( W ref ) ( 1 ) ⟩ = exp ( − 2 π i [ a N + μ ∗ ( N ∪ M W ref ) ] ) .
The local factor is conjugated by the first slot, and the same slot reverses the filling N N N in the closed double. This is (5) , with the local and spectral terms at the same specified connection. It uses the normalized section and reference framework of Tachikawa–Zhang [3 equations (67), (70)–(73)] .
Changing W r e f W_{\mathrm{ref}} W ref adds a closed framed trivial-gauge background. Its actual virtual operator, eta phase and finite tangent transgression cancel, so the scalar is independent of that reference choice. On an oriented differential String cylinder with trivial gauges, the same zero coefficient and Y 8 = 0 Y_8=0 Y 8 = 0 make (18) agree with the canonical reference transport. The scalar is therefore preserved by such cylinders and by oriented differential String isomorphisms, with their Spin and differential identifications retained.
Writing a = a N a=a_N a = a N and μ = μ ∗ ( Z ) \mu=\mu_*(Z) μ = μ ∗ ( Z ) , both minus signs in ϕ = − a − μ \phi=-a-\mu ϕ = − a − μ are fixed by this pairing. Changing only the eta sign would yield − a + μ -a+\mu − a + μ , whereas complex conjugation changes both terms to a + μ a+\mu a + μ . Those operations coincide only if 2 a ∈ Z 2a\in\mathbb{Z} 2 a ∈ Z . No such condition is imposed here, and the disk local term is generally nonzero. The sign dictionary agrees with the negative Green–Schwarz sign in the published scalar [3 equations (71)–(73)] .
For the sphere bundles in this paper, let B B B be a closed connected spin seven-manifold with specified Spin structure and metric. A principal right K = Sp ( 1 ) K=\operatorname{Sp}(1) K = Sp ( 1 ) bundle P → B P\to B P → B with connection gives V = P × K H V=P\times_K\mathbb{H} V = P × K H through the left multiplication action of K K K ; right multiplication by i i i defines its underlying complex rank-two bundle U U U . In the real quaternionic orientation ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) , put c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) and choose an integral differential trivialization h ˇ \check{h} h ˇ of λ ( T B ) + c ˇ \lambda(TB)+\check{c} λ ( T B ) + c ˇ , with three-form h h h . Check accents denote the corresponding differential cocycles and trivializations. On checked cochains, d d d denotes the full cochain differential, whose form component is the exterior derivative. The resulting relations are
c = c 2 ( U ) , e ( V R ) = λ ( V R ) = − c , d h ˇ = λ ˇ ( T B ) + c ˇ , d h = λ ( T B ) + c . c=c_2(U),\qquad e(V_{\mathbb{R}})=\lambda(V_{\mathbb{R}})=-c,\qquad d\check{h}=\check{\lambda}(TB)+\check{c},\qquad dh=\lambda(TB)+c. c = c 2 ( U ) , e ( V R ) = λ ( V R ) = − c , d h ˇ = λ ˇ ( T B ) + c ˇ , d h = λ ( T B ) + c .
Orient N = D ( V ) N=D(V) N = D ( V ) by the base followed by this fibre and M = S ( V ) M=S(V) M = S ( V ) as its outward boundary, and write π : N → B \pi:N\to B π : N → B for the disk projection. The gauge bundles are E = 2 π ∗ V R ⊕ R 8 E=2\pi^*V_{\mathbb{R}}\oplus\mathbb{R}^8 E = 2 π ∗ V R ⊕ R 8 and E ′ = R 16 E'=\mathbb{R}^{16} E ′ = R 16 , with their actual Spin lifts. For the connection A A A on the pulled-back principal bundle chosen in the disk construction, put C = c 2 ( π ∗ P , A ) C=c_2(\pi^*P,A) C = c 2 ( π ∗ P , A ) . Then n = − 2 C n=-2C n = − 2 C , p 2 ( E ) = 6 C 2 p_2(E)=6C^2 p 2 ( E ) = 6 C 2 , and (7) specializes to X 4 = λ ( A T ) + 2 C X_4 = \lambda(A_T) + 2C X 4 = λ ( A T ) + 2 C , Y 8 = − C 2 / 3 Y_8 = -C^2/3 Y 8 = − C 2 /3 . § 3 constructs the full induced boundary String datum and evaluates a N a_N a N at these connections. § 4 then fills the actual closed spin/gauge double to evaluate μ ∗ \mu_* μ ∗ . Their combination measures the retained integral background data in (5) .
Quaternionic sphere bundles and the local counterterm
The response of § 2 requires a filling with more structure than a spin manifold and a gauge connection. Its three-form must refine an integral trivialization of X 4 X_4 X 4 , and its boundary must carry the specified String datum and gauge frames. A quaternionic disk bundle constructs all of these data together. It also reduces the local contribution to an integral over the seven-dimensional base. That contribution is generally nonzero; it is needed when the local and spectral terms are combined in § 4 . We follow Li and Sun’s differential disk construction [7 Section 3] , retaining the connections and boundary identifications that determine this integral. Its evaluation uses exact identities after integration over the disk fibre; the corresponding pointwise curvature identities on the total space are not needed.
The disk geometry and its framed gauge boundary
Let B B B be a closed connected spin seven-manifold with specified Spin structure and metric. Take a principal right K = Sp ( 1 ) K = \operatorname{Sp}(1) K = Sp ( 1 ) bundle P P P with connection a a a , and form the quaternionic line
V = P × K H , ( p k , x ) ∼ ( p , k x ) . V = P \times_K \mathbb{H}, \qquad(pk,x) \sim(p,kx). V = P × K H , ( p k , x ) ∼ ( p , k x ) .
The group acts by left multiplication. Right multiplication by i i i gives the underlying complex rank-two bundle U U U . With quaternionic real orientation ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) , the complex orientation is ( 1 , i , j , − k ) (1,i,j,-k) ( 1 , i , j , − k ) , so our characteristic conventions are
c = c 2 ( U ) , e ( V R ) = − c , λ ( V R ) = − c , λ = p 1 2 . (19) c = c_2(U), \qquad e(V_{\mathbb{R}}) = -c, \qquad\lambda(V_{\mathbb{R}}) = -c, \qquad\lambda= \frac{p_1}{2}. \tag*{(19)} c = c 2 ( U ) , e ( V R ) = − c , λ ( V R ) = − c , λ = 2 p 1 . (19)
This orientation difference fixes the sign of the relative gauge number below. We use the same symbols for characteristic classes and their Chern–Weil forms when the connection is specified; statements about differential cocycles retain their check accents.
The base input is an integral differential trivialization h ˇ \check{h} h ˇ :
d h ˇ = λ ˇ ( T B ) + c ˇ , d h = λ ( T B ) + c . (20) d\check{h} = \check{\lambda}(TB) + \check{c}, \qquad dh = \lambda(TB) + c. \tag*{(20)} d h ˇ = λ ˇ ( T B ) + c ˇ , d h = λ ( T B ) + c . (20)
Here h h h is its three-form component, and c ˇ \check{c} c ˇ uses the connection a a a . The first equation includes the integral nullhomotopy and its geometric refinement. The second records only the curvature equation. In particular, solving the second equation over the reals does not supply the first. The construction applies to these specified ( B , P , a , h ˇ ) (B,P,a,\check{h}) ( B , P , a , h ˇ ) data; it does not assert that every quaternionic line on every spin base admits such a trivialization. No rational-cohomology vanishing or restriction on the fundamental group is imposed in this construction.
Set N = D ( V ) N=D(V) N = D ( V ) and M = S ( V ) = ∂ N M=S(V)=\partial N M = S ( V ) = ∂ N , of dimensions eleven and ten, and write π : N → B \pi:N\to B π : N → B for the bundle projection. Orient N N N by the base followed by the quaternionic real fibre, and M M M by the outward-boundary convention. The horizontal distribution of a splits the tangent bundle:
T N ≅ π ∗ ( T B ⊕ V R ) . (21) TN\cong\pi^*(TB\oplus V_{\mathbb R}). \tag*{(21)} T N ≅ π ∗ ( T B ⊕ V R ) . (21)
The specified Spin structure on B B B and the quaternionic representation give its product Spin structure. The actual gauge bundles on N N N are
E = 2 π ∗ V R ⊕ R 8 , E ′ = R 16 . (22) E=2\pi^*V_{\mathbb R}\oplus\mathbb R^8,\qquad E'=\mathbb R^{16}. \tag*{(22)} E = 2 π ∗ V R ⊕ R 8 , E ′ = R 16 . (22)
Writing L u L_u L u for left multiplication by u u u , the representation u ↦ diag ( L u , L u , I 8 ) u\mapsto\operatorname{diag}(L_u,L_u,I_8) u ↦ diag ( L u , L u , I 8 ) of K K K in SO ( 16 ) \operatorname{SO}(16) SO ( 16 ) has a unique identity-normalized lift to Spin ( 16 ) \operatorname{Spin}(16) Spin ( 16 ) because K K K is simply connected. Thus these are genuine common-cover Spin gauge bundles, rather than a formal choice of Pontryagin classes. In integral cohomology,
λ ( T N ) − λ ( E ) − λ ( E ′ ) = π ∗ ( λ ( T B ) + c ) . (23) \lambda(TN)-\lambda(E)-\lambda(E')=\pi^*(\lambda(TB)+c). \tag*{(23)} λ ( T N ) − λ ( E ) − λ ( E ′ ) = π ∗ ( λ ( T B ) + c ) . (23)
Equation (20) supplies its integral trivialization. No special splitting of T B TB T B is used.
The remaining task is to choose connections that give the required boundary identification. Write Q = π ∗ P Q=\pi^*P Q = π ∗ P . In a local base gauge, write a nonzero disk coordinate as x = r u x=ru x = r u , ∣ u ∣ = 1 |u|=1 ∣ u ∣ = 1 , and choose a smooth radial function f f f with f ( r ) = r 2 g ( r 2 ) f(r)=r^2g(r^2) f ( r ) = r 2 g ( r 2 ) near the centre and f = 1 f=1 f = 1 throughout a boundary collar. On Q Q Q put
A = ( 1 − f ) a − f d u u − 1 , C = c 2 ( Q , A ) . (24) A=(1-f)a-f\,du\,u^{-1},\qquad C=c_2(Q,A). \tag*{(24)} A = ( 1 − f ) a − f d u u − 1 , C = c 2 ( Q , A ) . (24)
Although u u u is undefined at the centre, r 2 d u u − 1 = Im ( d x x ˉ ) r^2du\,u^{-1}=\operatorname{Im}(dx\,\bar{x}) r 2 d u u − 1 = Im ( d x x ˉ ) makes the connection smooth there. Under p ↦ p k p\mapsto pk p ↦ p k , the coordinate changes by u ↦ k − 1 u u\mapsto k^{-1}u u ↦ k − 1 u and the expression transforms as A ↦ k − 1 A k + k − 1 d k A\mapsto k^{-1}Ak+k^{-1}dk A ↦ k − 1 A k + k − 1 d k ; it therefore defines a connection on the stated bundle.
At a unit vector v ∈ V b v\in V_b v ∈ V b , the map q ↦ v q q\mapsto vq q ↦ v q identifies H \mathbb H H with the pulled-back line. These maps give the tautological quaternionic frame τ \tau τ on M M M , extended radially to the collar. In that frame,
A τ = u − 1 A u + u − 1 d u = ( 1 − f ) ( u − 1 a u + u − 1 d u ) = 0 on the collar . (25) A^\tau=u^{-1}Au+u^{-1}du=(1-f)(u^{-1}au+u^{-1}du)=0\quad\text{on the collar}. \tag*{(25)} A τ = u − 1 A u + u − 1 d u = ( 1 − f ) ( u − 1 a u + u − 1 d u ) = 0 on the collar . (25)
The fixed representation lifts carry this frame to the gauge bundles in (22) . Their boundary connections are trivial in those specified Spin frames. Flatness alone would not have specified the frame used to identify the boundary with a reference.
Finite transgression and the induced String datum
Choose an invariant vertical metric d r 2 + ℓ ( r ) 2 g S 3 dr^2+\ell(r)^2g_{S^3} d r 2 + ℓ ( r ) 2 g S 3 , where g S 3 g_{S^3} g S 3 is the unit round sphere metric and the radial scale ℓ ( r ) \ell(r) ℓ ( r ) makes the metric Euclidean near the centre and product near the boundary. Its vertical natural connection ∇ n a t v \nabla^v_{\mathrm{nat}} ∇ nat v is the vertical Levi–Civita connection on vertical arguments and the associated connection under horizontal K K K -transport. Using the radial positive isometry on the same stabilized Spin bundle, compare
A ∗ = ∇ n a t v ⊕ π ∗ A B L C , A 0 = π ∗ ( A B L C ⊕ a V ) , (26) A_*=\nabla^v_{\mathrm{nat}}\oplus\pi^*A_B^{\mathrm{LC}},\qquad A_0=\pi^*(A_B^{\mathrm{LC}}\oplus a_V), \tag*{(26)} A ∗ = ∇ nat v ⊕ π ∗ A B LC , A 0 = π ∗ ( A B LC ⊕ a V ) , (26)
T q = T c 2 ( π ∗ a , A ) , K 3 = T λ ( A 0 , A ∗ ) . (27) T_q=T_{c_2}(\pi^*a,A),\qquad K_3=T_\lambda(A_0,A_*). \tag*{(27)} T q = T c 2 ( π ∗ a , A ) , K 3 = T λ ( A 0 , A ∗ ) . (27)
Here A B L C A_{B}^{\mathrm{LC}} A B LC is the Levi–Civita connection of the base metric, a V a_{V} a V is the connection induced on V R V_{\mathbb{R}} V R , and for a characteristic polynomial P P P the transgression convention is d T P ( A 0 , A 1 ) = P ( A 1 ) − P ( A 0 ) dT_{P}(A_{0},A_{1})=P(A_{1})-P(A_{0}) d T P ( A 0 , A 1 ) = P ( A 1 ) − P ( A 0 ) . The connection A ∗ A_{*} A ∗ preserves the outward normal and is product on the collar, as required for subsequent gluing. The intermediate A 0 A_{0} A 0 is used to calculate the finite change in characteristic forms; it is not the gluing connection. Their identical base summands cancel in K 3 K_{3} K 3 . What remains is the transgression of a fixed vertical equivariant model. This matching will also control its fibre integral.
To lift these form equations to integral differential data, fix the quaternionic and Whitney-sum characteristic morphisms. For Spin bundles F 1 , F 2 F_{1},F_{2} F 1 , F 2 they read
λ ~ ( V R ) ≃ − c ~ 2 ( U ) , λ ~ ( F 1 ⊕ F 2 ) ≃ λ ~ ( F 1 ) + λ ~ ( F 2 ) . (28) \widetilde{\lambda}(V_{\mathbb{R}})\simeq-\widetilde{c}_{2}(U),\qquad\widetilde{\lambda}(F_{1}\oplus F_{2})\simeq\widetilde{\lambda}(F_{1})+\widetilde{\lambda}(F_{2}). \tag*{(28)} λ ( V R ) ≃ − c 2 ( U ) , λ ( F 1 ⊕ F 2 ) ≃ λ ( F 1 ) + λ ( F 2 ) . (28)
The symbol ≃ \simeq ≃ denotes the chosen characteristic morphism, whose pullbacks will identify the cocycles throughout the construction. Their integral classes and curvature polynomials agree. At the universal level the possible flat differences lie in degree-three R / Z \mathbb{R}/\mathbb{Z} R / Z cohomology of the relevant products of the classifying spaces B S p i n B\mathrm{Spin} B Spin and B S p ( 1 ) B\mathrm{Sp}(1) B Sp ( 1 ) , which vanishes. Choose the unit morphism on framed trivial bundles, and retain the pullbacks and compositions in this construction. This specifies the morphisms being used; it does not make arbitrary morphisms on B B B unique. In the geometric model of differential trivializations, the induced identification depends on the actual morphism, as Redden explains [15 Proposition 3.17] .
Let q ~ = c ~ 2 ( Q , A ) \widetilde{q}=\widetilde{c}_{2}(Q,A) q = c 2 ( Q , A ) , q ~ 0 = π ∗ c ~ \widetilde{q}_{0}=\pi^{*}\widetilde{c} q 0 = π ∗ c , and λ ~ B = λ ~ ( T B , A B L C ) \widetilde{\lambda}_{B}=\widetilde{\lambda}(TB,A_{B}^{\mathrm{LC}}) λ B = λ ( T B , A B LC ) . Interpolating between the connections along an interval, with the specified endpoint morphisms, gives geometric transgression cochains T ~ , K ~ \widetilde{T},\widetilde{K} T , K satisfying
d T ~ = q ~ − q ~ 0 , d K ~ = λ ~ ( A ∗ ) − π ∗ ( λ ~ B − c ~ ) . (29) d\widetilde{T}=\widetilde{q}-\widetilde{q}_{0},\qquad d\widetilde{K}=\widetilde{\lambda}(A_{*})-\pi^{*}(\widetilde{\lambda}_{B}-\widetilde{c}). \tag*{(29)} d T = q − q 0 , d K = λ ( A ∗ ) − π ∗ ( λ B − c ) . (29)
They include integral and real cochain components as well as the forms T q , K 3 T_{q},K_{3} T q , K 3 . These equations are not isomorphisms between differential characters of unequal curvature. Appendix § 13 describes the cochain information needed here and in the relative pairing.
The full disk datum is
H ~ N = π ∗ h ~ + 2 T ~ + K ~ , H N = π ∗ h + 2 T q + K 3 , (30) \widetilde{H}_{N}=\pi^{*}\widetilde{h}+2\widetilde{T}+\widetilde{K},\qquad H_{N}=\pi^{*}h+2T_{q}+K_{3}, \tag*{(30)} H N = π ∗ h + 2 T + K , H N = π ∗ h + 2 T q + K 3 , (30)
d H ~ N = λ ~ ( A ∗ ) + 2 q ~ , d H N = λ ( A ∗ ) + 2 C . (31) d\widetilde{H}_{N}=\widetilde{\lambda}(A_{*})+2\widetilde{q},\qquad dH_{N}=\lambda(A_{*})+2C. \tag*{(31)} d H N = λ ( A ∗ ) + 2 q , d H N = λ ( A ∗ ) + 2 C . (31)
Indeed, the full cochain differential is
d H ~ N = π ∗ ( λ ~ B + c ~ ) + 2 ( q ~ − π ∗ c ~ ) + λ ~ ( A ∗ ) − π ∗ ( λ ~ B − c ~ ) = λ ~ ( A ∗ ) + 2 q ~ . (32) \begin{aligned}
d\widetilde{H}_{N}
&=\pi^{*}(\widetilde{\lambda}_{B}+\widetilde{c})+2(\widetilde{q}-\pi^{*}\widetilde{c})+\widetilde{\lambda}(A_{*})-\pi^{*}(\widetilde{\lambda}_{B}-\widetilde{c})\\
&=\widetilde{\lambda}(A_{*})+2\widetilde{q}. \tag*{(32)}
\end{aligned} d H N = π ∗ ( λ B + c ) + 2 ( q − π ∗ c ) + λ ( A ∗ ) − π ∗ ( λ B − c ) = λ ( A ∗ ) + 2 q . (32)
Both base contributions cancel with their indicated coefficients. This is the differential refinement of (23) at the selected connections. The finite transgressions carry the base trivialization to the connections and boundary geometry used in the normalized response. Although their form contributions to the local integral will vanish below, the integral datum still retains their cochains and endpoint morphisms.
The tautological boundary frame supplies a specified cochain morphism γ ~ T \widetilde{\gamma}_{T} γ T with d γ ~ T = q ~ ∣ M d\widetilde{\gamma}_{T}=\widetilde{q}|_{M} d γ T = q ∣ M . Its three-form component is zero in the normalized framed flat model. Its remaining components still specify the identification with the trivial gauge datum. Splitting off the outward normal by the product Spin identification gives
s ˇ M = H ˇ N ∣ M − 2 γ ˇ τ , d s ˇ M = λ ˇ ( T M ) . (33) \check{s}_M=\check{H}_N|_M-2\check{\gamma}_\tau,\qquad d\check{s}_M=\check{\lambda}(TM). \tag*{(33)} s ˇ M = H ˇ N ∣ M − 2 γ ˇ τ , d s ˇ M = λ ˇ ( T M ) . (33)
The factor two removes the two quaternionic gauge summands in their specified frames. What remains is a geometric trivialization of the integral tangent class λ ( T M ) \lambda(TM) λ ( T M ) , namely the differential String datum on the sphere bundle. An independent integral shift of s ˇ M \check{s}_M s ˇ M changes this input, even when the curvature equation has the same appearance; an actual differential isomorphism is needed to identify two such data.
The exact fibre integral of the local term
We can now evaluate the local term at these connections. The full gauge polynomial of § 2 uses n = λ ( E ) n=\lambda(E) n = λ ( E ) and n ′ = λ ( E ′ ) n'=\lambda(E') n ′ = λ ( E ′ ) :
Y 8 = n 2 + n n ′ + n ′ 2 − p 2 ( E ) − p 2 ( E ′ ) 6 . Y_8=\frac{n^2+nn'+n'^{2}-p_2(E)-p_2(E')}{6}. Y 8 = 6 n 2 + n n ′ + n ′ 2 − p 2 ( E ) − p 2 ( E ′ ) .
For one quaternionic line, the total Pontryagin form is 1 − 2 C + C 2 1-2C+C^2 1 − 2 C + C 2 . Two copies consequently give n = − 2 C n=-2C n = − 2 C and p 2 ( E ) = 6 C 2 p_2(E)=6C^2 p 2 ( E ) = 6 C 2 , whereas the framed trivial second gauge factor has n ′ = p 2 ( E ′ ) = 0 n'=p_2(E')=0 n ′ = p 2 ( E ′ ) = 0 . Substitution yields
X 4 = λ ( A ∗ ) + 2 C , Y 8 = 4 C 2 − 6 C 2 6 = − 1 3 C 2 . (34) X_4=\lambda(A_*)+2C,\qquad Y_8=\frac{4C^2-6C^2}{6}=-\frac{1}{3}C^2. \tag*{(34)} X 4 = λ ( A ∗ ) + 2 C , Y 8 = 6 4 C 2 − 6 C 2 = − 3 1 C 2 . (34)
The second Pontryagin term is essential to this coefficient. The local integral is therefore
a N = ∫ N H N Y 8 = − 1 3 ∫ N ( π ∗ h + 2 T q + K 3 ) C 2 . (35) a_N=\int_N H_NY_8=-\frac{1}{3}\int_N(\pi^*h+2T_q+K_3)C^2. \tag*{(35)} a N = ∫ N H N Y 8 = − 3 1 ∫ N ( π ∗ h + 2 T q + K 3 ) C 2 . (35)
Its evaluation requires identities for the selected forms, not merely their cohomology classes: h h h need not be closed, and the transgression terms are part of the specified differential background. There is a related distinction between C C C and π ∗ c \pi^*c π ∗ c . They represent the same ordinary characteristic class on the disk, but C C C is the curvature form of the radial connection, with vertical components and a fixed trivialization on the boundary. A pullback form π ∗ c \pi^*c π ∗ c would have no vertical components; moreover, c 2 c^2 c 2 is an eight-form on a seven-dimensional base. Substituting it for C C C inside (35) would erase the relative fibre information that the calculation needs. The connection and frame specify precisely how that information enters.
The universal K K K -equivariant quaternionic disk supplies those identities. Its radial connection is − f d u u − 1 -f\,du\,u^{-1} − f d u u − 1 , with moment correction ( 1 − f ) (1-f) ( 1 − f ) times the Lie algebra variable. Cartan association with the base connection a a a gives exactly (24) . In relative equivariant cohomology, its second Chern class is C K = − U T h C_K=-U_{\mathrm{Th}} C K = − U Th , where U T h U_{\mathrm{Th}} U Th is the oriented Thom class. The tautological section has positive Euler number, while (19) gives e = − c e=-c e = − c . In this universal calculation c c c denotes the class in H 4 ( B K ) H^4(BK) H 4 ( B K ) whose associated Chern–Weil form is c c c on B B B . The Thom relation and fibre normalization are
U T h 2 = − c U T h , ∫ D 4 U T h = 1 , ∫ D 4 C K 2 = − c . U_{\mathrm{Th}}^2=-cU_{\mathrm{Th}},\qquad\int_{D^4}U_{\mathrm{Th}}=1,\qquad\int_{D^4}C_K^2=-c. U Th 2 = − c U Th , ∫ D 4 U Th = 1 , ∫ D 4 C K 2 = − c .
After fibre integration the universal expression lies in the Cartan complex of a point: the K K K -invariant polynomials on its Lie algebra, with each polynomial variable assigned degree two. Association replaces these variables by the curvature of a a a . The differential of this target complex is zero, so there are no nonzero exact polynomials; the cohomology calculation determines the integrated polynomial itself. Cartan association commutes with oriented integration. Writing π ∗ \pi_* π ∗ for integration over the oriented disk fibre, we obtain the exact form identity
π ∗ C 2 = − c . (36) \pi_* C^2=-c. \tag*{(36)} π ∗ C 2 = − c . (36)
Exactness here refers to the fibre-integrated expressions. The relative cohomology calculation alone would not establish a corresponding pointwise form identity on N N N ; the zero differential in the universal target is what makes the integrated identity exact.
The same target eliminates the other two contributions. T q T_q T q and K 3 K_3 K 3 arise from degree-three equivariant forms on the fixed vertical model; for K 3 K_3 K 3 , this uses the cancellation of the base connections in (26) . Multiplying by C K 2 C_K^2 C K 2 and integrating over a four-dimensional disk leaves total degree seven. The Cartan complex of a point has no odd-degree part, so after association
π ∗ ( T q C 2 ) = 0 , π ∗ ( K 3 C 2 ) = 0 (37) \pi_*(T_q C^2)=0,\qquad\pi_*(K_3 C^2)=0 \tag*{(37)} π ∗ ( T q C 2 ) = 0 , π ∗ ( K 3 C 2 ) = 0 (37)
as forms. The gauge curvature is zero throughout the framed collar; the relative calculation has no additional boundary contribution. The projection formula now gives
π ∗ ( ( π ∗ h + 2 T q + K 3 ) C 2 ) = − h ∧ c . \pi_*\left((\pi^*h+2T_q+K_3)C^2\right)=-h\wedge c. π ∗ ( ( π ∗ h + 2 T q + K 3 ) C 2 ) = − h ∧ c .
Here the two degree-seven transgression pushforwards vanish, while the pullback of h h h multiplies π ∗ C 2 = − c \pi_*C^2=-c π ∗ C 2 = − c . Substitution into (35) gives
a N = 1 3 ∫ B h ∧ c . (38) a_N=\frac{1}{3}\int_B h\wedge c. \tag*{(38)} a N = 3 1 ∫ B h ∧ c . (38)
The eleven-dimensional integral has reduced to a seven-dimensional one, with degree-three and degree-four forms. This is a local counterterm in the specified anomaly response, entering the definition of ϕ \phi ϕ with sign − a N -a_N − a N . It is not by itself the closed response. A special representative such as the h = 0 h=0 h = 0 datum used in the S p ( 2 ) Sp(2) S p ( 2 ) calibration makes it vanish; a general base datum does not.
The outward fibre period
The disk also fixes a useful boundary period. Restrict the chosen tangent connection A ∗ A_* A ∗ to a quaternionic disk fibre. Its base summand has no fibre curvature. For the warped vertical metric, each curvature entry is proportional to either d r ∧ θ a dr\wedge\theta_a d r ∧ θ a or θ a ∧ θ b \theta_a\wedge\theta_b θ a ∧ θ b , where the θ a \theta_a θ a are angular coframe forms. The self-wedges of these entries vanish, and hence tr ( F ∧ F ) = 0 \operatorname{tr}(F\wedge F)=0 tr ( F ∧ F ) = 0 . The tangent λ \lambda λ form restricts to zero on the fibre. The relative gauge number in the tautological frame is ∫ D 4 C = − 1 \int_{D^4}C=-1 ∫ D 4 C = − 1 . Stokes’ theorem applied to (31) therefore gives
∫ ∂ D 4 H N = ∫ D 4 ( λ ( A ∗ ) + 2 C ) = − 2. (39) \int_{\partial D^4}H_N=\int_{D^4}\left(\lambda(A_*)+2C\right)=-2. \tag*{(39)} ∫ ∂ D 4 H N = ∫ D 4 ( λ ( A ∗ ) + 2 C ) = − 2. (39)
Here the fibre sphere is oriented as the boundary of the quaternionic disk. This is a real period at the specified connection. The original right-framed positive-unit datum on S p ( 2 ) \mathrm{Sp}(2) Sp ( 2 ) requires its separate inversion and differential transport; its period + 1 +1 + 1 is not the generic period of this construction.
Section 7 7 7 will use the distinction between fibre period and full String datum when changing backgrounds. For the immediate calculation, the disk has now provided the actual boundary geometry and the local contribution (38) . What remains is the spectral term of the closed manifold N ‾ ∪ M W r e f \overline{N} \cup_M W_{\mathrm{ref}} N ∪ M W ref , formed by gluing the orientation-reversed disk to the reference filling of Section 2 2 2 . The next section computes it with an auxiliary twelve-dimensional spin and gauge filling. That filling serves a different purpose from the differential H \mathcal{H} H -filling constructed here.
The closed eta phase from a twelve-dimensional filling
The disk calculation in Section 3 3 3 determines the local contribution a N a_N a N , but the normalized response also contains the closed eta invariant of Z = N ‾ ∪ M W r e f Z = \overline{N} \cup_M W_{\mathrm{ref}} Z = N ∪ M W ref . The gauge connection is trivial in the specified frame on the common boundary M = S ( V ) M = S(V) M = S ( V ) ; it is not trivial throughout the closed manifold. Its extension across the two halves carries information that the local boundary curvature cannot record. We evaluate this closed term by constructing a compact spin twelve-manifold with boundary Z Z Z and extending its actual gauge bundles and connections. Ordinary Atiyah–Patodi–Singer (APS) theory then relates the eta invariant to a bulk integral, with the real parity fixing the factor of one half.
The filling is auxiliary geometry for this calculation. Its existence does not assert an additional spacetime or an extension of the differential String trivialization. This distinction allows the closed term to be nonzero. The following construction realizes the mixed double as an associated four-sphere bundle and extends it over a spin filling of ( B , V ) (B,V) ( B , V ) [7 §4] . The resulting APS integral requires no shrinking-fibre limit.
Doubling the disk with its framed reference
Choose W r e f W_{\mathrm{ref}} W ref to be a second spin disk bundle with the same product tangent geometry on M M M , equipped with trivial principal gauge bundles and their constant Spin frames. On the other half, the tautological frame of the pulled-back quaternionic line supplies the boundary identification. Gluing these specified frames produces
Z = N ‾ ∪ M W r e f = P × K F , F = S ( H ⊕ R ) , K = S p ( 1 ) . (40) Z = \overline{N} \cup_M W_{\mathrm{ref}} = P \times_K F,\qquad F = S(\mathbb{H} \oplus\mathbb{R}),\qquad K = \mathrm{Sp}(1). \tag*{(40)} Z = N ∪ M W ref = P × K F , F = S ( H ⊕ R ) , K = Sp ( 1 ) . (40)
The term “mixed double” refers to the different gauge data on the two halves. Although their underlying disk geometry agrees, one half has the quaternionic gauge bundle and the other has the framed trivial reference. An ordinary double with identical gauges would compute a different closed background.
Orient F F F as the outward sphere of H ⊕ R \mathbb{H} \oplus\mathbb{R} H ⊕ R , with H \mathbb{H} H listed first. In the two disk charts, the lower chart has orientation − H -\mathbb{H} − H and the upper chart has orientation + H +\mathbb{H} + H . The gauge line on the lower chart transforms by left multiplication by K K K ; on the upper chart its chosen coefficient frame has trivial K K K -action. The transition on the equator is the unit quaternion u u u . Denote the resulting K K K -equivariant quaternionic line by L F L_F L F . Its fibre number is
∫ F c 2 ( L F ) = 1. (41) \int_F c_2(L_F)=1. \tag*{(41)} ∫ F c 2 ( L F ) = 1. (41)
Indeed, the relative disk number in the tautological frame was − 1 -1 − 1 . That disk appears with reversed orientation in N ‾ \overline{N} N , giving + 1 +1 + 1 after gluing to the trivial upper half. This is the orientation needed in the subsequent pushforwards.
The coefficient is the actual equivariant tautological line on H P 1 \mathbb{H}P^1 H P 1 , rather than a line chosen solely to have this characteristic number. The projective charts exhibit the identification. The action of diag ( u , 1 ) \operatorname{diag}(u,1) diag ( u , 1 ) on [ 1 : z ] [1:z] [ 1 : z ] sends z z z to z u − 1 zu^{-1} z u − 1 and acts by u u u on the coefficient frame. With the coordinate x = z ‾ x=\overline{z} x = z , the base action becomes x ↦ u x x\mapsto ux x ↦ ux . On [ w : 1 ] [w:1] [ w : 1 ] , the action is w ↦ u w w\mapsto uw w ↦ u w , while the coefficient frame is invariant. These are precisely the lower and upper chart actions of the glued line. Retaining them fixes how the fibre construction will associate to the bundle P P P on B B B .
The disk connection also glues as a connection, including the collar. In a local base gauge its form is A = ( 1 − f ) a − f d u u − 1 A=(1-f)a-f\,du\,u^{-1} A = ( 1 − f ) a − f d u u − 1 , where f = 1 f=1 f = 1 throughout a boundary collar. In the tautological frame, A τ = ( 1 − f ) ( u − 1 a u + u − 1 d u ) A^\tau=(1-f)(u^{-1}au+u^{-1}du) A τ = ( 1 − f ) ( u − 1 a u + u − 1 d u ) , so it vanishes on that collar. It therefore extends by zero in the specified upper frame, smoothly across the equator. No change of gauge connection at the seam is hidden in (40) .
The Spin identification is equally concrete. The equivariant stable tangent isomorphism
T F ⊕ R ≅ F × ( H ⊕ R ) (42) TF\oplus\mathbb{R}\cong F\times(\mathbb{H}\oplus\mathbb{R}) \tag*{(42)} T F ⊕ R ≅ F × ( H ⊕ R ) (42)
uses the action u ↦ diag ( L u , 1 ) u\mapsto\operatorname{diag}(L_u,1) u ↦ diag ( L u , 1 ) on the right-hand side, where L u L_u L u denotes left multiplication by u u u on H \mathbb{H} H . Since K K K is simply connected, this representation lifts to Spin ( 5 ) \operatorname{Spin}(5) Spin ( 5 ) . Removing the outward normal recovers the Spin lifts used on the disk charts. Combining the vertical lift with the specified Spin structure of B B B gives the Spin structure already present on the mixed double. Thus the associated description fixes the manifold, the twisting bundle and the Spin data of the operator whose eta invariant enters the response.
A spin and gauge filling of the base pair
To fill Z Z Z , first fill the pair ( B , V ) (B,V) ( B , V ) by an eight-dimensional spin manifold with an actual quaternionic line. The relevant bordism group is ordinary spin bordism with an SU ( 2 ) \operatorname{SU}(2) SU ( 2 ) bundle. Write B SU ( 2 ) B\operatorname{SU}(2) B SU ( 2 ) for the classifying space for that bundle and Ω q Spin \Omega_q^{\operatorname{Spin}} Ω q Spin for the bordism group of closed spin q q q -manifolds. The degree-seven vanishing follows directly from the Atiyah–Hirzebruch spectral sequence, whose second page is
E p , q 2 = H p ( B SU ( 2 ) ; Ω q Spin ) , B SU ( 2 ) = H P ∞ . (43) E^2_{p,q}=H_p(B\operatorname{SU}(2);\Omega_q^{\operatorname{Spin}}),\qquad B\operatorname{SU}(2)=\mathbb{H}P^\infty. \tag*{(43)} E p , q 2 = H p ( B SU ( 2 ) ; Ω q Spin ) , B SU ( 2 ) = H P ∞ . (43)
The cells of H P ∞ \mathbb{H}P^\infty H P ∞ occur only in dimensions divisible by four. Its cellular chain complex therefore has no adjacent nonzero groups, even with the coefficient group Ω q Spin \Omega_q^{\operatorname{Spin}} Ω q Spin . In total degree seven, the only possibly nonzero positions are ( p , q ) = ( 0 , 7 ) (p,q)=(0,7) ( p , q ) = ( 0 , 7 ) and ( 4 , 3 ) (4,3) ( 4 , 3 ) , but Ω 7 S p i n = Ω 3 S p i n = 0 \Omega^{\mathrm{Spin}}_{7}=\Omega^{\mathrm{Spin}}_{3}=0 Ω 7 Spin = Ω 3 Spin = 0 . Positions with negative q q q do not contribute. The entire degree-seven diagonal therefore vanishes already on the second page, giving
Ω 7 S p i n ( B S U ( 2 ) ) = 0. (44) \Omega^{\mathrm{Spin}}_{7}(BSU(2))=0. \tag*{(44)} Ω 7 Spin ( B S U ( 2 )) = 0. (44)
This is the ordinary spin/SU(2) result discussed by Davighi–Lohitsiri [18 §4.1, Figure 4] . It extends the classifying map of the actual bundle, not merely its Chern class. Neither this argument nor its application imposes a condition on π 1 ( B ) \pi_{1}(B) π 1 ( B ) .
Choose such a compact spin eight-manifold W W W with ∂ W = B \partial W=B ∂ W = B and quaternionic line V ~ \widetilde{V} V restricting to V V V . Fix the boundary Spin and bundle isomorphisms. Extend the metric and connection with product collars, and write
c W = c 2 ( U ~ ) , λ W = λ ( T W ) , λ = p 1 / 2. (45) c_{W}=c_{2}(\widetilde{U}),\qquad\lambda_{W}=\lambda(TW),\qquad\lambda=p_{1}/2. \tag*{(45)} c W = c 2 ( U ) , λ W = λ ( T W ) , λ = p 1 /2. (45)
Here U ~ \widetilde{U} U is the underlying complex rank-two bundle of V ~ \widetilde{V} V , with the same right-complex convention as U U U on B B B . These symbols denote the characteristic forms when used in integrals; their integral classes have the same normalization. The boundary trivialization h ~ \widetilde{h} h of λ ~ ( T B ) + c ~ \widetilde{\lambda}(TB)+\widetilde{c} λ ( T B ) + c is still fixed. We do not demand that it extend over W W W . The bordism result concerns spin and gauge data, whereas extending that trivialization would be a stronger requirement. Its boundary form h h h will remain in the answer.
Let P ~ \widetilde{P} P be the principal K K K bundle of V ~ \widetilde{V} V . Associate the same equivariant fibre and coefficient as above:
Q = P ~ × K F p ⟶ W , L = P ~ × K L F , Q=\widetilde{P}\mathbin{\times_{K}}F^{p}\longrightarrow W,\qquad L=\widetilde{P}\mathbin{\times_{K}}L_{F}, Q = P × K F p ⟶ W , L = P × K L F ,
E = 2 L R ⊕ R 8 , E ′ = R 16 . (46) \mathcal{E}=2\mathcal{L}_{\mathbb{R}}\oplus\mathbb{R}^{8},\qquad\mathcal{E}'=\mathbb{R}^{16}. \tag*{(46)} E = 2 L R ⊕ R 8 , E ′ = R 16 . (46)
The vertical Spin lift combines with that on W W W . The homomorphism K → S O ( 16 ) K\to SO(16) K → S O ( 16 ) acting on the two real quaternionic summands has the specified lift to Spin ( 16 ) \operatorname{Spin}(16) Spin ( 16 ) , so these are actual principal gauge bundles on the Spin ( 16 ) × Spin ( 16 ) \operatorname{Spin}(16)\times\operatorname{Spin}(16) Spin ( 16 ) × Spin ( 16 ) cover. On the base collar every datum is pulled back from B B B . Consequently ∂ Q = Z \partial Q=Z ∂ Q = Z with the required Spin, gauge and connection identifications, rather than only with isomorphic characteristic classes. Orient Q Q Q by W W W followed by F F F . Its boundary is then oriented by B B B followed by F F F ; the even fibre dimension introduces no additional sign.
For the characteristic-form calculation, use the invariant fibre metric, its natural vertical connection, and the split tangent connection
A Q , ∗ = ∇ n a t v ⊕ p ∗ A W L C . (47) A_{Q,*}=\nabla^{v}_{\mathrm{nat}}\oplus p^{*}A^{\mathrm{LC}}_{W}. \tag*{(47)} A Q , ∗ = ∇ nat v ⊕ p ∗ A W LC . (47)
It restricts to the specified stable tangent connection on Z Z Z . The analytic APS operator will first be taken with a product-collar Levi–Civita connection; we account below for the finite difference between that connection and (47) .
Exact fibre integration of the anomaly polynomial
Set C = c 2 ( L ) C=c_{2}(\mathcal{L}) C = c 2 ( L ) and ℓ = λ ( T v Q ) \ell=\lambda(T^{v}Q) ℓ = λ ( T v Q ) , where T v Q T^{v}Q T v Q is the tangent bundle along the fibres of p p p . The exact integrations needed for the density can be determined on the universal K K K -equivariant four-sphere. Write C K C_K C K and ℓ K \ell_K ℓ K for its equivariant gauge and vertical tangent characteristic classes, respectively, and c c c for the universal base class. Then
C K 2 = c C K , ℓ K = − c , ∫ F C K = 1 (48) C_K^2=cC_K,\qquad\ell_K=-c,\qquad\int_F C_K=1 \tag*{(48)} C K 2 = c C K , ℓ K = − c , ∫ F C K = 1 (48)
in equivariant cohomology. The first relation retains the chart orientation: the lower chart has orientation − H -\mathbb{H} − H and Euler class + c +c + c . Its compactly supported Chern class is the positive Thom class, normalized to integrate to one along the fibre; its square is c c c times that class. The gauge class restricts to c c c at the lower pole and to zero at the upper pole. The second relation follows from the stable tangent identification (42) and λ ( V R ) = − c \lambda(V_{\mathbb{R}})=-c λ ( V R ) = − c .
These cohomology relations alone would not justify pointwise substitution in an index density. What makes the present calculation exact at the level of forms is the target of the fibre integration. The equivariant integrals lie in the Cartan complex of a point: the invariant polynomials on the Lie algebra of K K K , with polynomial degree counted twice in the total degree. Its differential is zero. Thus the classes of the integrated expressions determine the invariant polynomials themselves, with no exact-form ambiguity in this target. The Cartan association map substitutes the curvature of the extended connection for the Lie algebra variable and commutes with fibre integration. Associating this universal model with the extended connection on P ~ \widetilde{P} P gives exact integrated forms. With p ∗ p_* p ∗ denoting oriented integration over the four-dimensional fibre, they are
p ∗ C = 1 , p ∗ C 2 = c W , p ∗ C 3 = c W 2 , p ∗ ( ℓ C 2 ) = − c W 2 . (49) p_*C=1,\qquad p_*C^2=c_W,\qquad p_*C^3=c_W^2,\qquad p_*(\ell C^2)=-c_W^2. \tag*{(49)} p ∗ C = 1 , p ∗ C 2 = c W , p ∗ C 3 = c W 2 , p ∗ ( ℓ C 2 ) = − c W 2 . (49)
The first identity fixes the normalization, the next two integrate the gauge terms, and the last keeps the vertical tangent contribution. They are equalities of forms at the associated connections. We have not asserted the pointwise curvature equalities C 2 = p ∗ c W C C^2=p^*c_WC C 2 = p ∗ c W C or ℓ = − p ∗ c W \ell=-p^*c_W ℓ = − p ∗ c W on Q Q Q .
For the actual gauge bundles in (46) , the same full anomaly polynomial used on the disk gives
I 12 , ∗ = − 1 3 ( p ∗ λ W + ℓ + 2 C ) C 2 . (50) I_{12,*}=-\frac{1}{3}(p^*\lambda_W+\ell+2C)C^2. \tag*{(50)} I 12 , ∗ = − 3 1 ( p ∗ λ W + ℓ + 2 C ) C 2 . (50)
The tangent term splits as p ∗ λ W + ℓ p^*\lambda_W+\ell p ∗ λ W + ℓ . For one real quaternionic line, p 1 = − 2 C p_1=-2C p 1 = − 2 C and p 2 = C 2 p_2=C^2 p 2 = C 2 . The Whitney product for its two copies is ( 1 − 2 C + C 2 ) 2 (1-2C+C^2)^2 ( 1 − 2 C + C 2 ) 2 , so λ ( E ) = − 2 C \lambda(\mathcal{E})=-2C λ ( E ) = − 2 C and p 2 ( E ) = 6 C 2 p_2(\mathcal{E})=6C^2 p 2 ( E ) = 6 C 2 ; the second gauge is trivial. Hence X 4 = p ∗ λ W + ℓ + 2 C X_4=p^*\lambda_W+\ell+2C X 4 = p ∗ λ W + ℓ + 2 C and Y 8 = − C 2 / 3 Y_8=-C^2/3 Y 8 = − C 2 /3 , with the gauge p 2 p_2 p 2 contribution included. The projection formula and (49) now yield
p ∗ I 12 , ∗ = − 1 3 ( λ W c W − c W 2 + 2 c W 2 ) = − 1 3 c W ( λ W + c W ) . (51) \begin{aligned}
p_*I_{12,*}=-\frac{1}{3}(\lambda_Wc_W-c_W^2+2c_W^2) \\
&=-\frac{1}{3}c_W(\lambda_W+c_W). \tag*{(51)}
\end{aligned} p ∗ I 12 , ∗ = − 3 1 ( λ W c W − c W 2 + 2 c W 2 ) = − 3 1 c W ( λ W + c W ) . (51)
The vertical tangent term cancels one of the two gauge contributions to c W 2 c_W^2 c W 2 . A remaining gauge term combines with λ W c W \lambda_Wc_W λ W c W ; the density does not vanish. The derivation is polynomial, so it uses no division by c W c_W c W and remains applicable when that form vanishes or its cohomology class is torsion.
This also displays why Q \mathcal{Q} Q is a spin/gauge filling rather than an H \mathcal{H} H -filling. On a fibre, the pullback base term has zero integral, and the vertical class ℓ K = − c \ell_K=-c ℓ K = − c contributes zero after restriction to the ordinary fibre. Therefore
∫ F X 4 = 2 ∫ F C = 2. (52) \int_F X_4=2\int_F C=2. \tag*{(52)} ∫ F X 4 = 2 ∫ F C = 2. (52)
A global trivialization of X 4 X_4 X 4 cannot exist there. In particular, one cannot replace the bulk polynomial by d ( H 3 Y 8 ) d(H_3Y_8) d ( H 3 Y 8 ) on Q \mathcal{Q} Q and invoke closed differential H \mathcal{H} H cancellation. Ordinary APS needs the Spin structure and twisting bundles that we have constructed, and does not require that stronger trivialization.
Real parity and the finite tangent correction
Use the real virtual coefficient
Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , Δ C = Δ R ⊗ R C . (53) \Delta_{\mathbb{R}}=\mathcal{E}\otimes_{\mathbb{R}}\mathcal{E}'-S_{\mathbb{R}}^{+}(\mathcal{E})-S_{\mathbb{R}}^{+}(\mathcal{E}'),\qquad\Delta_{\mathbb{C}}=\Delta_{\mathbb{R}}\otimes_{\mathbb{R}}\mathbb{C}. \tag*{(53)} Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , Δ C = Δ R ⊗ R C . (53)
The half-spin representations are real. With the ordinary Levi–Civita Dirac operator on Z Z Z , combine the eta representatives of these actual bundles before reduction, and use
ξ ( D ) = η A P S ( D ) + dim C ker D 2 , μ L C ( Z ) = − 1 2 ξ ( D Z , Δ C ) ( m o d Z ) . (54) \xi(D)=\frac{\eta_{\mathrm{APS}}(D)+\dim_{\mathbb{C}}\ker D}{2},\qquad\mu_{\mathrm{LC}}(Z)=-\frac{1}{2}\xi(D_{Z,\Delta_{\mathbb{C}}})\pmod{\mathbb{Z}}. \tag*{(54)} ξ ( D ) = 2 η APS ( D ) + dim C ker D , μ LC ( Z ) = − 2 1 ξ ( D Z , Δ C ) ( mod Z ) . (54)
The kernel term is part of the reduced invariant. No invertibility assumption is needed, and the real half is taken on the representatives in this formula.
Apply the twisted APS index theorem on the compact manifold Q \mathcal{Q} Q with a product-collar Levi–Civita metric and the extended gauge connection [8 Theorem (4.2), formula (4.3)] . It gives
ξ ( D Z , Δ C ) = ∫ Q [ A ^ ( T Q ) ch ( Δ C ) ] 12 − Ind A P S D Δ C + . (55) \xi(D_{Z,\Delta_{\mathbb{C}}})=\int_{\mathcal{Q}}\left[\widehat{A}(T\mathcal{Q})\operatorname{ch}(\Delta_{\mathbb{C}})\right]_{12}-\operatorname{Ind}_{\mathrm{APS}}D_{\Delta_{\mathbb{C}}}^{+}. \tag*{(55)} ξ ( D Z , Δ C ) = ∫ Q [ A ( T Q ) ch ( Δ C ) ] 12 − Ind APS D Δ C + . (55)
Because 12 ≡ 4 ( m o d 8 ) 12\equiv4\pmod{8} 12 ≡ 4 ( mod 8 ) , the chiral Spin representations with real twists have parallel quaternionic structures. They are compatible with the Dirac operator and the APS projections, including the zero eigenspaces. They preserve the kernel of the chiral APS problem and the kernel of its adjoint, whose boundary condition is complementary. Both kernels therefore have even complex dimension, so their difference, the complex APS index, is even. The statement persists under subtraction for the real virtual coefficient. This is the parity that removes the possible half-index ambiguity; the corresponding antilinear symmetry and APS normalization are described in [17 (2.47)–(2.52), (3.8)–(3.9)] and in Appendix § 14 .
Multiplying (55) by − 1 / 2 -1/2 − 1/2 , with I 12 = − [ A ^ ch ( Δ C ) ] 12 / 2 I_{12}=-[\widehat{A}\operatorname{ch}(\Delta_{\mathbb{C}})]_{12}/2 I 12 = − [ A ch ( Δ C ) ] 12 /2 , gives
μ L C ( Z ) = ∫ Q I 12 , L C + 1 2 Ind A P S D Δ C + = ∫ Q I 12 , L C ( m o d Z ) . (56) \mu_{\mathrm{LC}}(Z)=\int_{\mathcal{Q}}I_{12,\mathrm{LC}}+\frac{1}{2}\operatorname{Ind}_{\mathrm{APS}}D_{\Delta_{\mathbb{C}}}^{+}=\int_{\mathcal{Q}}I_{12,\mathrm{LC}}\pmod{\mathbb{Z}}. \tag*{(56)} μ LC ( Z ) = ∫ Q I 12 , LC + 2 1 Ind APS D Δ C + = ∫ Q I 12 , LC ( mod Z ) . (56)
Without the evenness of the actual index, this congruence would not follow from complex APS with this normalization. In particular, an arbitrary square root of a complex exponentiated identity cannot replace the real-coefficient argument.
The response uses μ ∗ \mu_* μ ∗ at the stable connection (47) . Keep the gauge connection fixed, and let T I ( A L C , A Q , ∗ ) T_I(A_{\mathrm{LC}}, A_{Q,*}) T I ( A LC , A Q , ∗ ) be the tangent Chern–Simons transgression with
d T I ( A L C , A Q , ∗ ) = I 12 , ∗ − I 12 , L C . (57) \mathrm{d}T_I(A_{\mathrm{LC}}, A_{Q,*}) = I_{12,*} - I_{12,\mathrm{LC}}. \tag*{(57)} d T I ( A LC , A Q , ∗ ) = I 12 , ∗ − I 12 , LC . (57)
Stokes’ theorem changes the bulk integral by the boundary integral of this form. By the differential-index convention of § 2 , that same finite boundary integral is added to μ L C \mu_{\mathrm{LC}} μ LC to define μ ∗ \mu_* μ ∗ . The two changes therefore agree:
μ ∗ ( Z ) = μ L C ( Z ) + ∫ Z T I ( A L C , A Q , ∗ ) = ∫ Q I 12 , ∗ = − 1 3 ∫ W c W ( λ W + c W ) ( m o d Z ) . (58) \begin{aligned}
\mu_*(Z) &= \mu_{\mathrm{LC}}(Z) + \int_Z T_I(A_{\mathrm{LC}}, A_{Q,*}) \\
&= \int_Q I_{12,*} = -\frac{1}{3}\int_W c_W(\lambda_W+c_W) \pmod{\mathbb{Z}}. \tag*{(58)}
\end{aligned} μ ∗ ( Z ) = μ LC ( Z ) + ∫ Z T I ( A LC , A Q , ∗ ) = ∫ Q I 12 , ∗ = − 3 1 ∫ W c W ( λ W + c W ) ( mod Z ) . (58)
All spectral operators in this derivation are ordinary Levi–Civita operators. The stable connection enters through its finite transgression, not through a proposed spectrum for a torsionful operator. Thus the form calculation and the analytic boundary value refer to the same specified data.
Combining the disk and closed contributions
The disk result was a N = ∫ B h c / 3 a_N=\int_B hc/3 a N = ∫ B h c /3 . With the normalized scalar ϕ = − a N − μ ∗ ( Z ) \phi=-a_N-\mu_*(Z) ϕ = − a N − μ ∗ ( Z ) , equation (58) gives
ϕ ( S ( V ) , s ˘ M ) = 1 3 ( ∫ W c W ( λ W + c W ) − ∫ B h c ) ( m o d Z ) . (59) \phi(S(V),\breve{s}_M)=\frac{1}{3}\left(\int_W c_W(\lambda_W+c_W)-\int_B hc\right)\pmod{\mathbb{Z}}. \tag*{(59)} ϕ ( S ( V ) , s ˘ M ) = 3 1 ( ∫ W c W ( λ W + c W ) − ∫ B h c ) ( mod Z ) . (59)
The positive bulk term comes from the minus sign on the closed eta term; the boundary correction retains the negative sign on the disk term. Dropping a N a_N a N would change this response for general boundary data. Neither the polynomial cancellation above nor the existence of a spin/gauge filling sets ∫ B h c \int_B hc ∫ B h c to zero.
The evaluation (59) [7 §4] applies to the sphere-bundle background already specified: a closed spin seven-manifold with the actual quaternionic line and the integral differential trivialization of λ + c \lambda+c λ + c . No rational-cohomology vanishing was used. The existence of the actual filling is essential: it permits APS to evaluate the closed eta phase, including information that a match of local polynomials alone would not fix. The separate family calculation in Appendix § 16 provides another route; it is not an assumption of this evaluation.
The parentheses in (59) have the form of a relative integral: the bulk eight-form has boundary primitive h c hc h c because d h = λ + c \mathrm{d}h=\lambda+c d h = λ + c and d c = 0 \mathrm{d}c=0 d c = 0 . Its integrality, however, requires the retained integral nullhomotopy in h ˘ \breve{h} h ˘ , not just that curvature equation. This nullhomotopy supplies a relative class z ∈ H 4 ( W , B ; Z ) z\in H^4(W,B;\mathbb{Z}) z ∈ H 4 ( W , B ; Z ) whose absolute image is λ ( T W ) + c 2 ( U ~ ) \lambda(TW)+c_2(\widetilde{U}) λ ( T W ) + c 2 ( U ) . § 5 identifies the integral combination with K = ⟨ c W ⌣ z , [ W , B ] ⟩ K=\langle c_W\smile z,[W,B]\rangle K = ⟨ c W ⌣ z , [ W , B ]⟩ , the pairing with the relative fundamental class of the oriented filling. That identification converts the geometric evaluation into ϕ = K / 3 \phi=K/3 ϕ = K /3 and specifies its filling ambiguity. It is also the starting point for the Crowley–Goette comparison on its stated rational-cohomology domain.
The relative characteristic number and the factor of eight
The APS calculation in § 4 has reduced the ten-dimensional response to a bulk integral on an eight-manifold and a boundary integral on its seven-dimensional base. Their combination is
ϕ ( S ( V ) , s ˇ M ) = 1 3 { ∫ W c W ( λ W + c W ) − ∫ B h ∧ c } in R / Z . (60) \phi(S(V),\check{s}_M)=\frac{1}{3}\left\{\int_W c_W(\lambda_W+c_W)-\int_B h\wedge c\right\}\ \text{in }\mathbb{R}/\mathbb{Z}. \tag*{(60)} ϕ ( S ( V ) , s ˇ M ) = 3 1 { ∫ W c W ( λ W + c W ) − ∫ B h ∧ c } in R / Z . (60)
Here ∂ W = B \partial W=B ∂ W = B , the actual quaternionic line V ~ \widetilde{V} V extends V V V , and c W = c 2 ( U ~ ) c_W=c_2(\widetilde{U}) c W = c 2 ( U ) , where U ~ \widetilde{U} U is its underlying complex rank-two bundle. The convention is λ W = p 1 ( T W ) / 2 \lambda_W=p_1(TW)/2 λ W = p 1 ( T W ) /2 ; the extension includes the specified boundary identifications of the Spin structures, bundles, and connections. The three-form h h h is the form component of the integral differential trivialization h ˇ \check{h} h ˇ on B B B ; it satisfies d h = λ ( T B ) + c dh=\lambda(TB)+c d h = λ ( T B ) + c . The orientation of B B B is the boundary orientation of W W W , compatible with the base-then-fibre and outward-boundary orientations used to construct S ( V ) S(V) S ( V ) .
Neither integral in (60) is by itself the retained topological datum. The bulk integral uses an auxiliary extension, and the boundary term is generally nonzero. The full integral nullhomotopy in h ˇ \check{h} h ˇ ties these two pieces together: their difference is an integral relative characteristic number. This explains why the response is quantized even though its calculation passes through differential forms. Li and Sun identify this difference with an integral relative pairing and use it to compare the response with the Crowley–Goette invariant [7 §5] .
The boundary trivialization fixes the relative class
Let j : H 4 ( W , B ; Z ) → H 4 ( W ; Z ) j:H^4(W,B;\mathbb{Z})\to H^4(W;\mathbb{Z}) j : H 4 ( W , B ; Z ) → H 4 ( W ; Z ) forget the boundary trivialization. The specified integral nullhomotopy determines
z ∈ H 4 ( W , B ; Z ) , j z = λ ( T W ) + c 2 ( U ~ ) . (61) z\in H^4(W,B;\mathbb{Z}),\qquad jz=\lambda(TW)+c_2(\widetilde{U}). \tag*{(61)} z ∈ H 4 ( W , B ; Z ) , j z = λ ( T W ) + c 2 ( U ) . (61)
This is a relative class: it contains the bulk characteristic class together with its chosen trivialization at the boundary. Choosing only an absolute class whose boundary restriction vanishes would omit the information specifying that trivialization. Likewise, solving the real equation for h h h would not determine z z z .
One can see the relation between the integral and differential data directly. Write the bulk differential cocycle for λ + c \lambda+c λ + c as ( x Z , x R , X ) (x_{\mathbb{Z}},x_{\mathbb{R}},X) ( x Z , x R , X ) , where x Z x_{\mathbb{Z}} x Z is an integral four-cochain, x R x_{\mathbb{R}} x R is a real three-cochain, and the curvature four-form is X = λ W + c W X=\lambda_W+c_W X = λ W + c W . Its boundary geometric trivialization is ( b , k , h ) (b,k,h) ( b , k , h ) , with integral three-cochain b b b , real two-cochain k k k , and three-form h h h . These components are the differential-cocycle model described in Appendix § 13 . Let i : B ↪ W i:B\hookrightarrow W i : B ↪ W be the inclusion and δ \delta δ the singular cochain coboundary. With I \mathcal{I} I denoting integration of forms on singular chains, the trivialization equations are
δ b = x Z ∣ B , I h − b − δ k = x R ∣ B , d h = X ∣ B . (62) \delta b=x_{\mathbb{Z}}|_B,\qquad\mathcal{I}h-b-\delta k=x_{\mathbb{R}}|_B,\qquad dh=X|_B. \tag*{(62)} δ b = x Z ∣ B , I h − b − δ k = x R ∣ B , d h = X ∣ B . (62)
The integral cocycle ( x Z , b ) (x_{\mathbb{Z}},b) ( x Z , b ) represents z z z . In the relative complex, whose differential is d r e l ( u , v ) = ( d u , i ∗ u − d v ) d_{\mathrm{rel}}(u,v)=(du,i^*u-dv) d rel ( u , v ) = ( d u , i ∗ u − d v ) , these equations give
( I X , I h ) − ( x Z , b ) = d r e l ( x R , − k ) . (63) (\mathcal{I}X,\mathcal{I}h)-(x_{\mathbb{Z}},b)=d_{\mathrm{rel}}(x_{\mathbb{R}},-k). \tag*{(63)} ( I X , I h ) − ( x Z , b ) = d rel ( x R , − k ) . (63)
Thus the pair of forms ( X , h ) (X,h) ( X , h ) represents the real image of this particular integral class. The equation identifies the real representative without replacing the integral cocycle or its boundary morphism by form data.
Multiplying by the absolute class c 2 ( U ~ ) c_2(\widetilde{U}) c 2 ( U ) and evaluating on [ W , B ] [W,B] [ W , B ] yields
K : = ⟨ c 2 ( U ~ ) ⌣ z , [ W , B ] ⟩ = ∫ W c W ( λ W + c W ) − ∫ B h ∧ c ∈ Z . (64) \mathrm{K} := \left\langle c_2(\widetilde{U}) \smile z,[W,B]\right\rangle= \int_W c_W(\lambda_W+c_W)-\int_B h\wedge c\in\mathbb{Z}. \tag*{(64)} K := ⟨ c 2 ( U ) ⌣ z , [ W , B ] ⟩ = ∫ W c W ( λ W + c W ) − ∫ B h ∧ c ∈ Z . (64)
The minus sign is the relative-integration convention: a pair ( α , β ) (\alpha,\beta) ( α , β ) integrates as ∫ W α − ∫ B β \int_W\alpha-\int_B\beta ∫ W α − ∫ B β . At the differential-cochain level the product includes the homotopy between cup and wedge products. Redden’s product of a geometric trivialization with a differential cocycle has integral trivializing component b ⌣ c 2 ( U ~ ) b\smile c_2(\widetilde{U}) b ⌣ c 2 ( U ) and form component h ∧ c h\wedge c h ∧ c [15 §3.4, Proposition 3.23] . These are precisely the two components used in (64) ; Appendix § 13 gives the relative-complex conventions and product details. A freely chosen real primitive need not have this integral lift, so the curvature equation alone would not prove the displayed integrality.
Substitution into the evaluated response gives
ϕ ( S ( V ) , s ~ M ) = K 3 , O ( S ( V ) , s ~ M ) = exp ( 2 π i K 3 ) . (65) \phi(S(V),\widetilde{s}_M)=\frac{\mathrm{K}}{3},\qquad\mathcal{O}(S(V),\widetilde{s}_M)=\exp\left(2\pi i\frac{\mathrm{K}}{3}\right). \tag*{(65)} ϕ ( S ( V ) , s M ) = 3 K , O ( S ( V ) , s M ) = exp ( 2 π i 3 K ) . (65)
The number K \mathrm{K} K measures the relative characteristic pairing of the specified background; the construction does not identify it with a new physical charge. Its integrality implies that this response takes values among the third roots of unity. The exponentiation forgets all but the residue of K \mathrm{K} K modulo three, whereas the relative construction retains more information about the background. No rational-cohomology vanishing has entered this calculation.
Closed periods and relative evenness
A choice of filling gives an integer K \mathrm{K} K , but the complete fixed boundary data determine its residue modulo 24. To compare two fillings, glue one to the orientation reversal of the other, using the same boundary Spin and bundle identifications and the same integral nullhomotopy. The boundary cochains cancel. The resulting closed spin eight-manifold X X X carries an actual quaternionic line L L L , and the difference of the two relative pairings is
Δ K = ⟨ c 2 ( L ) ( λ ( T X ) + c 2 ( L ) ) , [ X ] ⟩ = 12 Ind D L − 2 ∈ 24 Z . (66) \Delta\mathrm{K}=\left\langle c_2(L)(\lambda(TX)+c_2(L)),[X]\right\rangle=12\operatorname{Ind}D_{L-2}\in24\mathbb{Z}. \tag*{(66)} Δ K = ⟨ c 2 ( L ) ( λ ( T X ) + c 2 ( L )) , [ X ] ⟩ = 12 Ind D L − 2 ∈ 24 Z . (66)
Here L L L is regarded as a complex rank-two coefficient and 2 means two trivial complex coefficients. Since ch ( L − 2 ) = − c 2 ( L ) + c 2 ( L ) 2 / 12 + ⋯ \operatorname{ch}(L-2)=-c_2(L)+c_2(L)^2/12+\cdots ch ( L − 2 ) = − c 2 ( L ) + c 2 ( L ) 2 /12 + ⋯ and A ^ 4 ( T X ) = − λ ( T X ) / 12 \widehat{A}_4(TX)=-\lambda(TX)/12 A 4 ( T X ) = − λ ( T X ) /12 , the degree-eight density is c 2 ( L ) ( λ ( T X ) + c 2 ( L ) ) / 12 c_2(L)(\lambda(TX)+c_2(L))/12 c 2 ( L ) ( λ ( T X ) + c 2 ( L )) /12 . The last inclusion uses quaternionic index parity [9 (1.1)–(1.4)] . In eight dimensions the chiral Spin representations are real. Tensoring with a quaternionic line supplies a quaternionic structure commuting with the twisted Dirac operator, so its index is even. Subtracting twice the untwisted index preserves evenness. Consequently the closed characteristic number is a multiple of 24, rather than merely a multiple of 12.
This gluing compares spin/gauge extensions of the same boundary datum. It does not require h ~ \widetilde{h} h to extend as a trivialization over either eight-manifold. Such an extension would restrict the relative class and is not a premise of the response formula. The 24 Z 24\mathbb{Z} 24 Z ambiguity changes K / 3 K/3 K /3 by an integer, and therefore leaves the normalized scalar and its exponentiated response unchanged.
There is a separate parity statement for each relative pairing,
K = ⟨ z ⌣ j z − λ ( T W ) ⌣ z , [ W , B ] ⟩ ∈ 2 Z . (67) K=\langle z\smile jz-\lambda(TW)\smile z,[W,B]\rangle\in2\mathbb{Z}. \tag*{(67)} K = ⟨ z ⌣ j z − λ ( T W ) ⌣ z , [ W , B ]⟩ ∈ 2 Z . (67)
It follows from (5.2), since c 2 ( U ~ ) = j z − λ ( T W ) c_2(\widetilde{U})=jz-\lambda(TW) c 2 ( U ) = j z − λ ( T W ) . To see the parity without assuming an extension of the trivialization, extend the relative class z z z by zero to the spin double of W W W . For a degree-four mod-two class its square is the fourth Steenrod square S q 4 \mathrm{Sq}^4 Sq 4 ; the Wu formula on the double pairs this square with the fourth Wu class. On a spin manifold this class is w 4 = λ m o d 2 w_4=\lambda\ \mathrm{mod}\ 2 w 4 = λ mod 2 . The zero extension contributes nothing from the other half of the double, so the two terms in (5.8) agree modulo two. Their difference is even.
The two statements have different uses. Closed index parity controls the change under a different auxiliary filling; the relative Wu argument constrains the allowed integer for one filling with its specified boundary trivialization. In particular, relative evenness does not force the phase to vanish: an even integer can have a nonzero residue modulo three. This distinction will be useful for the six-dimensional normalization in § 8 .
An intrinsic seven-dimensional expression
The same answer can be obtained from an ordinary Dirac operator on B B B , with no explicit choice of W W W in the final expression. Use the actual complex bundle U U U underlying V V V , and the virtual complex coefficient 4 U − 8 4U-8 4 U − 8 . The numbers 4 4 4 and − 8 -8 − 8 are integer multiplicities of actual bundles; the virtual coefficient has rank zero. Its degree-eight index density is
ch ( U ~ ) = 2 − c W + c W 2 12 + ⋯ , A ^ 4 ( T W ) = − λ W 12 , \operatorname{ch}(\widetilde{U})=2-c_W+\frac{c_W^2}{12}+\cdots,\qquad\widehat{A}_4(TW)=-\frac{\lambda_W}{12}, ch ( U ) = 2 − c W + 12 c W 2 + ⋯ , A 4 ( T W ) = − 12 λ W ,
[ A ^ ( T W ) ch ( 4 U ~ − 8 ) ] 8 = 1 3 c W ( λ W + c W ) . (68) \left[\widehat{A}(TW)\operatorname{ch}(4\widetilde{U}-8)\right]_8=\frac{1}{3}c_W(\lambda_W+c_W). \tag*{(68)} [ A ( T W ) ch ( 4 U − 8 ) ] 8 = 3 1 c W ( λ W + c W ) . (68)
The cancellation of the rank term removes the pure gravitational degree-eight contribution. The remaining product contains exactly the characteristic density in (5.5), with its positive sign and factor 1 / 3 1/3 1/3 .
For product-collar geometry, complex APS for U ~ \widetilde{U} U and the trivial bundle, with these integer multiplicities, gives [8 ]
Ind D 4 U ~ − 8 = 1 3 ∫ W c W ( λ W + c W ) − ξ ( B , 4 U − 8 ) . \operatorname{Ind}D_{4\widetilde{U}-8}=\frac{1}{3}\int_W c_W(\lambda_W+c_W)-\xi(B,4U-8). Ind D 4 U − 8 = 3 1 ∫ W c W ( λ W + c W ) − ξ ( B , 4 U − 8 ) .
The index is an integer, so reduction modulo Z \mathbb{Z} Z removes it. Subtracting the boundary term therefore gives
ϕ ( S ( V ) , s ˘ M ) = ξ ( B , 4 U − 8 ) − 1 3 ∫ B h ∧ c = K 3 in R / Z . (69) \phi(S(V),\breve{s}_M)=\xi(B,4U-8)-\frac{1}{3}\int_B h\wedge c=\frac{K}{3}\quad\text{in }\mathbb{R}/\mathbb{Z}. \tag*{(69)} ϕ ( S ( V ) , s ˘ M ) = ξ ( B , 4 U − 8 ) − 3 1 ∫ B h ∧ c = 3 K in R / Z . (69)
The reduced invariant ξ ( D ) = ( η ( D ) + dim C ker D ) / 2 \xi(D)=(\eta(D)+\dim_{\mathbb{C}}\ker D)/2 ξ ( D ) = ( η ( D ) + dim C ker D ) /2 is extended additively to the virtual coefficient, including its zero modes. This is an intrinsic expression for the response already evaluated by the twelve-dimensional real APS calculation, not a replacement of its real eta normalization by an arbitrary half of a class in R / Z \mathbb{R}/\mathbb{Z} R / Z . The boundary correction still carries the differential trivialization. Omitting it would retain the spectral term while changing the scalar being computed.
The Crowley–Goette normalization
The comparison with the established secondary invariant requires an additional, precise restriction on the base:
H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0. (70) H^3(B;\mathbb{Q}) = H^4(B;\mathbb{Q}) = 0. \tag*{(70)} H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0. (70)
In the cited arXiv version, Crowley and Goette define their quaternionic-line t t t -invariant on closed spin manifolds with these rational-cohomology vanishings [9 Definition 1.4] . There is no simple-connectedness condition in this definition or in the filling comparison used here. The restriction has a concrete algebraic role. The sequence
H 3 ( B ; Q ) ⟶ H 4 ( W , B ; Q ) → j H 4 ( W ; Q ) ⟶ H 4 ( B ; Q ) (71) H^3(B;\mathbb{Q}) \longrightarrow H^4(W,B;\mathbb{Q}) \xrightarrow{j} H^4(W;\mathbb{Q}) \longrightarrow H^4(B;\mathbb{Q}) \tag*{(71)} H 3 ( B ; Q ) ⟶ H 4 ( W , B ; Q ) j H 4 ( W ; Q ) ⟶ H 4 ( B ; Q ) (71)
makes j j j an isomorphism. Hence c 2 ( U ~ ) c_2(\widetilde{U}) c 2 ( U ) has a unique rational relative lift, denoted c ˉ \bar{c} c ˉ . The vanishings remove both the obstruction to this lift and its rational ambiguity; they do not set integral torsion to zero.
For an actual spin/quaternionic filling, the dimension-seven extrinsic formula of Crowley–Goette is
t B ( V ) = − 1 2 ⟨ [ A ^ ( T W ) ch ′ ( U ~ ) ] 4 ⌣ c ˉ , [ W , B ] ⟩ ( m o d Z ) , 2 − ch ( U ~ ) = c W ch ′ ( U ~ ) . (72) t_B(V)=-\frac{1}{2}\left\langle[\widehat{A}(TW)\operatorname{ch}'(\widetilde{U})]_4\smile\bar{c},[W,B]\right\rangle\pmod{\mathbb{Z}},\qquad2-\operatorname{ch}(\widetilde{U})=c_W\operatorname{ch}'(\widetilde{U}). \tag*{(72)} t B ( V ) = − 2 1 ⟨ [ A ( T W ) ch ′ ( U ) ] 4 ⌣ c ˉ , [ W , B ] ⟩ ( mod Z ) , 2 − ch ( U ) = c W ch ′ ( U ) . (72)
Their general dimension 4 k − 1 4k-1 4 k − 1 formula contains 1 / a k + 1 1/a_{k+1} 1/ a k + 1 , with a j = 2 a_j=2 a j = 2 for odd j j j and a j = 1 a_j=1 a j = 1 for even j j j . Here k = 2 k=2 k = 2 and a 3 = 2 a_3=2 a 3 = 2 , giving the factor 1 / 2 1/2 1/2 in (72) [9 (1.3)–(1.6)] . Their Proposition 1.3 identifies this filling expression with the intrinsic invariant. The actual spin/quaternionic filling constructed in § 4 supplies its geometric premise; the cohomology restriction supplies the unique relative lift.
The degree-four calculation is short but fixes both sign and coefficient:
ch ′ ( U ~ ) = 1 − c W 12 + ⋯ , [ A ^ ( T W ) ch ′ ( U ~ ) ] 4 = − λ W + c W 12 . (73) \operatorname{ch}'(\widetilde{U})=1-\frac{c_W}{12}+\cdots,\qquad[\widehat{A}(TW)\operatorname{ch}'(\widetilde{U})]_4=-\frac{\lambda_W+c_W}{12}. \tag*{(73)} ch ′ ( U ) = 1 − 12 c W + ⋯ , [ A ( T W ) ch ′ ( U ) ] 4 = − 12 λ W + c W . (73)
The minus sign cancels the one in the extrinsic definition. If z Q z_{\mathbb{Q}} z Q denotes the rational image of the specified integral class z z z , uniqueness implies that it is the relative lift of λ W + c W \lambda_W+c_W λ W + c W . Compatibility and graded commutativity of relative and absolute products then give
t B ( V ) = 1 24 ⟨ ( λ W + c W ) ⌣ c ˉ , [ W , B ] ⟩ = 1 24 ⟨ c 2 ( U ~ ) ⌣ z Q , [ W , B ] ⟩ = K 24 ( m o d Z ) . (74) t_B(V)=\frac{1}{24}\left\langle(\lambda_W+c_W)\smile\bar{c},[W,B]\right\rangle
=\frac{1}{24}\left\langle c_2(\widetilde{U})\smile z_{\mathbb{Q}},[W,B]\right\rangle
=\frac{K}{24}\pmod{\mathbb{Z}}. \tag*{(74)} t B ( V ) = 24 1 ⟨ ( λ W + c W ) ⌣ c ˉ , [ W , B ] ⟩ = 24 1 ⟨ c 2 ( U ) ⌣ z Q , [ W , B ] ⟩ = 24 K ( mod Z ) . (74)
Thus the rational calculation evaluates the same pairing as the integral construction. It does not introduce a different choice of boundary data. At the level of forms, choose a primitive c ^ \widehat{c} c with d c ^ = c d\widehat{c}=c d c = c on B B B . The identity d ( h ∧ c ^ ) = c ^ ∧ ( λ + c ) − h ∧ c d(h\wedge\widehat{c})=\widehat{c}\wedge(\lambda+c)-h\wedge c d ( h ∧ c ) = c ∧ ( λ + c ) − h ∧ c then gives ∫ B c ^ ∧ ( λ + c ) = ∫ B h ∧ c \int_B\widehat{c}\wedge(\lambda+c)=\int_B h\wedge c ∫ B c ∧ ( λ + c ) = ∫ B h ∧ c by Stokes on closed B B B . Thus the two relative representatives have the same boundary correction, including its sign.
Combining (65) and (74) gives the comparison
ϕ ( S ( V ) , s ˇ M ) = 8 t B ( V ) , O ( S ( V ) , s ˇ M ) = exp ( 2 π i 8 t B ( V ) ) . (75) \phi(S(V),\check{s}_M)=8t_B(V),\qquad\mathcal{O}(S(V),\check{s}_M)=\exp(2\pi i\,8t_B(V)). \tag*{(75)} ϕ ( S ( V ) , s ˇ M ) = 8 t B ( V ) , O ( S ( V ) , s ˇ M ) = exp ( 2 π i 8 t B ( V )) . (75)
The factor eight compares two normalizations of one relative characteristic pairing: 1 / 3 1/3 1/3 for the Green–Schwarz scalar and 1 / 24 1/24 1/24 for t t t . Its denominator includes the quaternionic factor a 3 = 2 a_3=2 a 3 = 2 ; omitting that factor would change the answer. This is a comparison with Crowley–Goette’s existing invariant, rather than a new definition of t t t or a count of microscopic degrees of freedom.
The integral and comparison formulas retain different domains. Equation (65) applies to the full specified sphere-bundle backgrounds. Equation (75) additionally requires exactly (70) . Outside that domain the relative integer still computes the response, but it is not justified to call K / 24 K/24 K /24 the Crowley–Goette invariant. § 6 evaluates the actual line on the original S p ( 2 ) Sp(2) S p ( 2 ) example, which lies in the comparison domain. § 7 then uses the larger domain to show how changing the integral String datum can change the response on a fixed ten-manifold.
The calibrated response on S p ( 2 ) Sp(2) S p ( 2 )
The relative formula of § 5 becomes a concrete response on the group manifold S p ( 2 ) Sp(2) S p ( 2 ) . An actual quaternionic-line filling fixes the sign of the relative pairing and gives K = − 2 K=-2 K = − 2 , hence O = e 2 π i / 3 \mathcal{O}=e^{2\pi i/3} O = e 2 π i /3 . This is the known third-root phase obtained by Tachikawa–Zhang in their published calculation [3 §3.2, equation (63)] . Li and Sun’s actual-line evaluation and Spin frame identifications [7 §6] explain how the sphere-bundle calculation reaches the specified Lie-framed background. The pairing evaluates the induced left datum; finite differential transport then identifies its response with that of the original right-framed positive-unit datum. The connections, frames and integral differential trivialization are essential to this comparison.
Let e 1 , e 2 e_1,e_2 e 1 , e 2 be the standard coordinate vectors of H 2 \mathbb{H}^2 H 2 . Take B = S ( H 2 ) B=S(\mathbb{H}^2) B = S ( H 2 ) with its outward orientation and let P = S p ( 2 ) → B P=Sp(2)\to B P = S p ( 2 ) → B send p p p to p e 2 pe_2 p e 2 . Its right stabilizer is K = S p ( 1 ) K=Sp(1) K = S p ( 1 ) embedded as diag ( u , 1 ) \operatorname{diag}(u,1) diag ( u , 1 ) . The associated quaternionic line is V y = y ⊥ H V_y=y^{\perp_{\mathbb{H}}} V y = y ⊥ H , and hence
K ⟶ P ⟶ S 7 , N = D ( V ) , M = S ( V ) ≅ S p ( 2 ) . (76) K\longrightarrow P\longrightarrow S^7,\qquad N=D(V),\qquad M=S(V)\cong Sp(2). \tag*{(76)} K ⟶ P ⟶ S 7 , N = D ( V ) , M = S ( V ) ≅ S p ( 2 ) . (76)
The identification of the unit-vector bundle with P P P sends [ p , u ] [p,u] [ p , u ] to p diag ( u , 1 ) p\operatorname{diag}(u,1) p diag ( u , 1 ) . Thus the same principal bundle supplies the base gauge data and the total space on which the response is evaluated. Write π = π N : N → B \pi=\pi_N:N\to B π = π N : N → B for the disk projection. We orient N N N by the base followed by the real quaternionic fibre, and M M M by the outward boundary convention. The integer K K K remains distinct from the stabilizer group K K K .
The actual line and the Thom pairing
To evaluate K K K , the eight-dimensional filling must extend the actual line V V V with its boundary identification. A stable extension with a matching characteristic number would not specify the bundle needed by the associated spin and gauge filling. The useful filling here is the disk of the tautological line over H P 1 \mathbb{H}P^1 H P 1 , rather than a disk D 8 D^8 D 8 in which the gauge group has been enlarged.
Let L L L be that tautological line over H P 1 ≅ S 4 \mathbb{H}P^1 \cong S^4 H P 1 ≅ S 4 , with complement L ⊥ L^\perp L ⊥ in H P 1 × H 2 \mathbb{H}P^1 \times\mathbb{H}^2 H P 1 × H 2 , and let ρ \rho ρ be the disk projection. Set
W 0 = D ( L R ) , V ~ = ρ ∗ L ⊥ , ∂ W 0 = S ( L R ) ≅ S ( H 2 ) . (77) W_0 = D(L_{\mathbb{R}}), \qquad\widetilde{V} = \rho^*L^\perp, \qquad\partial W_0 = S(L_{\mathbb{R}}) \cong S(\mathbb{H}^2). \tag*{(77)} W 0 = D ( L R ) , V = ρ ∗ L ⊥ , ∂ W 0 = S ( L R ) ≅ S ( H 2 ) . (77)
The boundary map ( ℓ , y ) ↦ y (\ell,y) \mapsto y ( ℓ , y ) ↦ y identifies the fibre of V ~ \widetilde{V} V literally with y ⊥ H y^{\perp_{\mathbb{H}}} y ⊥ H . This is the eight-dimensional base filling used to calculate the relative integer; it is different from the eleven-dimensional disk N N N that defines the boundary response.
Orient the chart w ↦ [ 1 : w ] w \mapsto[1:w] w ↦ [ 1 : w ] by ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) , and normalize x ∈ H 4 ( H P 1 ; Z ) x \in H^4(\mathbb{H}P^1;\mathbb{Z}) x ∈ H 4 ( H P 1 ; Z ) by ∫ H P 1 x = 1 \int_{\mathbb{H}P^1} x=1 ∫ H P 1 x = 1 . The signs can be seen directly by projecting e 1 e_1 e 1 into L L L . The projection has one zero at [ 0 : 1 ] [0:1] [ 0 : 1 ] ; in the chart [ w : 1 ] [w:1] [ w : 1 ] its quaternionic coefficient is w ‾ / ( 1 + ∣ w ∣ 2 ) \overline{w}/(1+|w|^2) w / ( 1 + ∣ w ∣ 2 ) . Quaternionic conjugation reverses real orientation, while the transition between the charts is quaternionic inversion and preserves orientation in dimension four. Consequently
e ( L R ) = − x , c 2 ( L ) = x , c 2 ( L ⊥ ) = − x . (78) e(L_{\mathbb{R}})=-x,\qquad c_2(L)=x,\qquad c_2(L^\perp)=-x. \tag*{(78)} e ( L R ) = − x , c 2 ( L ) = x , c 2 ( L ⊥ ) = − x . (78)
These signs also retain the convention e ( V R ) = − c 2 ( U ) e(V_{\mathbb{R}})=-c_2(U) e ( V R ) = − c 2 ( U ) : the complex structure is right multiplication by i i i , whose complex orientation is opposite to ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) .
Orient W 0 W_0 W 0 by base then fibre. Near [ 1 : 0 ] [1:0] [ 1 : 0 ] , the blowdown to H 2 \mathbb{H}^2 H 2 is ( w , q ) ↦ ( q , w q ) (w,q)\mapsto(q,wq) ( w , q ) ↦ ( q , w q ) . At ( 0 , 1 ) (0,1) ( 0 , 1 ) its differential interchanges two four-dimensional blocks with sign ( − 1 ) 16 = + 1 (-1)^{16}=+1 ( − 1 ) 16 = + 1 and sends the outward radial normal to the outward radial normal. It therefore induces precisely the outward S 7 S^7 S 7 orientation used above. The boundary Spin structure is the unique one on S 7 S^7 S 7 . The tangent bundle of the disk splits as T W 0 ≅ ρ ∗ ( T S 4 ⊕ L R ) TW_0 \cong\rho^*(TS^4\oplus L_{\mathbb{R}}) T W 0 ≅ ρ ∗ ( T S 4 ⊕ L R ) . Since λ ( T S 4 ) = 0 \lambda(TS^4)=0 λ ( T S 4 ) = 0 and λ ( L R ) = − c 2 ( L ) \lambda(L_{\mathbb{R}})=-c_2(L) λ ( L R ) = − c 2 ( L ) , this gives
λ ( T W 0 ) = − ρ ∗ x , c W = c 2 ( V ~ ) = − ρ ∗ x . (79) \lambda(TW_0)=-\rho^*x,\qquad c_W=c_2(\widetilde{V})=-\rho^*x. \tag*{(79)} λ ( T W 0 ) = − ρ ∗ x , c W = c 2 ( V ) = − ρ ∗ x . (79)
Let u ∈ H 4 ( W 0 , S 7 ; Z ) u\in H^4(W_0,S^7;\mathbb{Z}) u ∈ H 4 ( W 0 , S 7 ; Z ) be the positive Thom class of L R L_{\mathbb{R}} L R , and let j j j denote the map from relative to absolute cohomology. The Thom identity gives j u = ρ ∗ e ( L R ) = − ρ ∗ x ju=\rho^*e(L_{\mathbb{R}})=-\rho^*x j u = ρ ∗ e ( L R ) = − ρ ∗ x . Both H 3 ( S 7 ; Z ) H^3(S^7;\mathbb{Z}) H 3 ( S 7 ; Z ) and H 4 ( S 7 ; Z ) H^4(S^7;\mathbb{Z}) H 4 ( S 7 ; Z ) vanish, so the exact sequence of the pair makes j j j an isomorphism here. Consequently the class specified by the boundary nullhomotopy is z = 2 u z=2u z = 2 u , since j z = λ ( T W 0 ) + c W = − 2 ρ ∗ x jz=\lambda(TW_0)+c_W=-2\rho^*x j z = λ ( T W 0 ) + c W = − 2 ρ ∗ x . Thom integration sends ρ ∗ x ⌣ u \rho^*x\smile u ρ ∗ x ⌣ u to x x x on H P 1 \mathbb{H}P^1 H P 1 , whose integral is one. The signed relative product therefore evaluates to
K = ⟨ ( − ρ ∗ x ) ⌣ 2 u , [ W 0 , S 7 ] ⟩ = − 2 , t S 7 ( V ) = − 1 12 , ϕ ( S ( V ) ) = 1 3 ( m o d Z ) . (80) K=\langle(-\rho^*x)\smile2u,[W_0,S^7]\rangle=-2,\qquad t_{S^7}(V)=-\frac{1}{12},\qquad\phi(S(V))=\frac{1}{3}\pmod{\mathbb{Z}}. \tag*{(80)} K = ⟨( − ρ ∗ x ) ⌣ 2 u , [ W 0 , S 7 ]⟩ = − 2 , t S 7 ( V ) = − 12 1 , ϕ ( S ( V )) = 3 1 ( mod Z ) . (80)
The last two values are in R / Z \mathbb{R}/\mathbb{Z} R / Z : the Crowley–Goette comparison gives t S 7 ( V ) = K / 24 t_{S^7}(V)=K/24 t S 7 ( V ) = K /24 , and the response formula gives ϕ = K / 3 = − 2 / 3 = 1 / 3 \phi=K/3=-2/3=1/3 ϕ = K /3 = − 2/3 = 1/3 modulo Z \mathbb{Z} Z . The comparison applies because S 7 S^7 S 7 has vanishing rational cohomology in degrees three and four. Product-collar extensions of the specified boundary geometry and connections preserve this relative evaluation; an extension of the base String trivialization over W 0 W_0 W 0 is not required.
The selected zero local term and finite transport
The relative number computes the response of the induced sphere-bundle datum. To identify it with the left Lie-framed datum s L s_L s L , one must also identify the actual Spin bundles, boundary frames and differential trivialization. Over S 7 S^7 S 7 there are special tangent splittings T B = V R ⊕ R 3 TB = V_{\mathbb{R}} \oplus\mathbb{R}^3 T B = V R ⊕ R 3 and T N ≅ π ∗ ( 2 V R ⊕ R 3 ) TN \cong\pi^*(2V_{\mathbb{R}} \oplus\mathbb{R}^3) T N ≅ π ∗ ( 2 V R ⊕ R 3 ) . They identify the actual gauge bundle E L = 2 Q H ⊕ R 8 E_L = 2Q_{\mathbb{H}} \oplus\mathbb{R}^8 E L = 2 Q H ⊕ R 8 with T N ⊕ R 5 TN \oplus\mathbb{R}^5 T N ⊕ R 5 , where Q = π ∗ P Q = \pi^*P Q = π ∗ P and Q H = π ∗ V R Q_{\mathbb{H}} = \pi^*V_{\mathbb{R}} Q H = π ∗ V R is its associated real quaternionic line. In the tautological boundary frame taken twice, the induced stable frame is the left Lie frame, with its fixed Spin lift. The diagonal Spin identification supplies the integral nullhomotopy. Appendix § 15 gives the frame comparison and its smooth product-collar connections.
This geometry admits a particularly useful representative. If a a a is the base principal connection and a V a_V a V its associated connection, choose
A B ∗ = a V ⊕ d R 3 , λ ( A B ∗ ) + c = 0. (81) A_B^* = a_V \oplus d_{\mathbb{R}^3}, \qquad\lambda(A_B^*) + c = 0. \tag*{(81)} A B ∗ = a V ⊕ d R 3 , λ ( A B ∗ ) + c = 0. (81)
The second equality is an equality of forms at this connection. The corresponding differential character is flat, and H 3 ( S 7 ; R / Z ) = 0 H^3(S^7;\mathbb{R}/\mathbb{Z}) = 0 H 3 ( S 7 ; R / Z ) = 0 makes it trivial. The integral base trivialization class is unique, so its geometric representative can be chosen with h = 0 h = 0 h = 0 , retaining the characteristic morphisms and integral cochains.
For the smooth gauge connection trivial in the tautological boundary frame, put C = c 2 ( Q , A ) C = c_2(Q,A) C = c 2 ( Q , A ) . The product tangent connection A ∗ A_* A ∗ uses the natural vertical connection of § 3 and base summand π ∗ A B ∗ \pi^*A_B^* π ∗ A B ∗ . For two copies of the quaternionic line, λ ( E L ) = − 2 C \lambda(E_L) = -2C λ ( E L ) = − 2 C and p 2 ( E L ) = 6 C 2 p_2(E_L) = 6C^2 p 2 ( E L ) = 6 C 2 . Thus the full polynomial of § 2 reduces to X 4 = λ ( A ∗ ) + 2 C X_4 = \lambda(A_*) + 2C X 4 = λ ( A ∗ ) + 2 C and Y 8 = ( 4 C 2 − 6 C 2 ) / 6 = − C 2 / 3 Y_8 = (4C^2 - 6C^2)/6 = -C^2/3 Y 8 = ( 4 C 2 − 6 C 2 ) /6 = − C 2 /3 . The three-form transgressions are T q = T c 2 ( π ∗ a , A ) T_q = T_{c_2}(\pi^*a,A) T q = T c 2 ( π ∗ a , A ) and K 3 = T λ ( A 0 , A ∗ ) K_3 = T_\lambda(A_0,A_*) K 3 = T λ ( A 0 , A ∗ ) , with A 0 = π ∗ ( A B ∗ ⊕ a V ) A_0 = \pi^*(A_B^* \oplus a_V) A 0 = π ∗ ( A B ∗ ⊕ a V ) . Thus both tangent connections have the same base summand. As in the disk calculation, the transgressions in H L = 2 T q + K 3 H_L = 2T_q + K_3 H L = 2 T q + K 3 have vanishing integrated odd-degree terms:
π N ∗ C 2 = − c , π N ∗ ( T q C 2 ) = π N ∗ ( K 3 C 2 ) = 0 , a N , ∗ = ∫ N H L Y 8 = 0. (82) \pi_{N*}C^2 = -c,\qquad\pi_{N*}(T_qC^2) = \pi_{N*}(K_3C^2) = 0,\qquad a_{N,*} = \int_N H_LY_8 = 0. \tag*{(82)} π N ∗ C 2 = − c , π N ∗ ( T q C 2 ) = π N ∗ ( K 3 C 2 ) = 0 , a N , ∗ = ∫ N H L Y 8 = 0. (82)
These are exact fibre-integrated form identities for the specified connections, obtained from the equivariant disk calculation. A cohomology-level vanishing alone would not justify dropping this local integral. The selected representative makes it zero; the general formula remains a N = 1 3 ∫ B h c a_N = \frac{1}{3}\int_B hc a N = 3 1 ∫ B h c .
The relation to the Levi–Civita base representative is a finite connection transport, not a replacement of the String datum. Let α = T λ ( A B L C , A B ∗ ) \alpha= T_\lambda(A_B^{\mathrm{LC}},A_B^*) α = T λ ( A B LC , A B ∗ ) and transport the same trivialization by h ∗ = h L C + α h_* = h_{\mathrm{LC}} + \alpha h ∗ = h LC + α , up to an exact form with its retained morphism. Write π : Z → B \pi: Z \to B π : Z → B for the projective model of the mixed double. The corresponding tangent transgression is − π ∗ α C 2 / 3 -\pi_* \alpha C^2/3 − π ∗ α C 2 /3 , and its closed-fibre pushforward uses π ∗ C 2 = c \pi_*C^2 = c π ∗ C 2 = c . In the following formula, the subscripts ∗ * ∗ , LC distinguish the two base connections; both closed terms include the finite tangent correction, using the real eta convention that includes kernels. Thus
a N , ∗ − a N , L C = 1 3 ∫ B α c , μ ∗ − μ L C = − 1 3 ∫ B α c . (83) a_{N,*} - a_{N,\mathrm{LC}} = \frac{1}{3}\int_B \alpha c,\qquad\mu_* - \mu_{\mathrm{LC}} = -\frac{1}{3}\int_B \alpha c. \tag*{(83)} a N , ∗ − a N , LC = 3 1 ∫ B α c , μ ∗ − μ LC = − 3 1 ∫ B α c . (83)
The two finite changes cancel in ϕ = − a N − μ \phi=-a_N-\mu ϕ = − a N − μ . Choosing the h = 0 h=0 h = 0 representative therefore puts the entire response in the corrected closed term. Changing back to the Levi–Civita base representative redistributes the value between the local and spectral terms, while preserving their combination. For the selected zero-local-term filling, μ ∗ = 2 / 3 \mu_* = 2/3 μ ∗ = 2/3 modulo Z \mathbb{Z} Z , and hence the left Lie-framed response is
ϕ ( P , s L ) = 1 3 ( m o d Z ) , O ( P , s L ) = e 2 π i / 3 . (84) \phi(P,s_L)=\frac{1}{3}\pmod{\mathbb{Z}},\qquad\mathcal{O}(P,s_L)=e^{2\pi i/3}. \tag*{(84)} ϕ ( P , s L ) = 3 1 ( mod Z ) , O ( P , s L ) = e 2 π i /3 . (84)
§ 10 uses this zero-counterterm filling.
The right-framed positive-unit datum
The outward stabilizer-fibre period of the sphere-bundle construction is − 2 -2 − 2 . The original right-framed positive-unit datum is reached by additional geometry specific to Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) . In particular, the right and left integral String structures differ over the identity map: s R − s L = 3 ν s_R-s_L=3\nu s R − s L = 3 ν , with ν ∈ H 3 ( P ; Z ) \nu\in H^3(P;\mathbb{Z}) ν ∈ H 3 ( P ; Z ) having positive unit period on the stabilizer S 3 S^3 S 3 . Agreement of a local polynomial or a bare flux number would not identify these data.
The right filling uses the actual bundle E R = T N ⊕ R 5 E_R=TN\oplus\mathbb{R}^5 E R = T N ⊕ R 5 , with a full Spin ( 16 ) \operatorname{Spin}(16) Spin ( 16 ) gauge connection extending the flat connection in the right boundary frame. It is not constrained to the two-quaternionic-block reduction used on the left. A product tangent connection A T A_T A T can be chosen with vanishing p 1 p_1 p 1 form on each disk fibre. Its diagonal transgression with the right gauge connection A G A_G A G supplies H R = T λ ( A G , A T ⊕ d R 5 ) H_R=T_\lambda(A_G,A_T\oplus d_{\mathbb{R}^5}) H R = T λ ( A G , A T ⊕ d R 5 ) , including the integral refinement. The relative gauge number in that boundary frame is − 1 -1 − 1 , so Stokes gives the real period
∫ S 3 H R = ∫ D 4 ( λ ( A T ) − λ ( A G ) ) = 0 − ( − 1 ) = 1. (85) \int_{S^3}H_R=\int_{D^4}\bigl(\lambda(A_T)-\lambda(A_G)\bigr)=0-(-1)=1. \tag*{(85)} ∫ S 3 H R = ∫ D 4 ( λ ( A T ) − λ ( A G ) ) = 0 − ( − 1 ) = 1. (85)
The full Spin polynomial and the specified boundary connections are essential here; a connection change does not change the relative class, and its boundary Chern–Simons term accounts for the change in the bulk integral.
Inversion ι ( p ) = p − 1 \iota(p)=p^{-1} ι ( p ) = p − 1 preserves orientation on the ten-dimensional group and satisfies ι ∗ s R = s L \iota^*s_R=s_L ι ∗ s R = s L . Its differential sends a left-invariant vector p X pX pX to − X p − 1 -Xp^{-1} − X p − 1 , a right-invariant vector with a constant minus sign. The constant transformation − I 10 -I_{10} − I 10 has positive determinant and can be joined to I 10 I_{10} I 10 while retaining the outward normal and the specified Spin lift. This identifies the framings after inversion; it does not identify them over the identity map of P P P . Pulling back the complete right metric, connections and refinement, then transporting them by differential String cylinders with trivial gauges, identifies its scalar with (84) . Appendix C describes the Spin lift and the required interval term in the differential cylinder; interpolation of endpoint three-forms alone would not supply this identification. On these cylinders the trivial gauges give Y 8 = Δ = 0 Y_8=\Delta=0 Y 8 = Δ = 0 , so the transport preserves the normalized response. Bare three-form periods need not remain fixed as the tangent connection changes. The result recovers the known third root on both specified data without asserting a positive-unit transport for an arbitrary sphere bundle.
Finally, reversing orientation together with the associated boundary, differential and reference data changes ϕ \phi ϕ to − ϕ -\phi − ϕ and O \mathcal{O} O to its complex conjugate. For this datum the reversed response is e − 2 π i / 3 e^{-2\pi i/3} e − 2 π i /3 . Independently resetting a reversed period to + 1 +1 + 1 would specify a different input. The dependence on integral String data, which this calibration makes concrete, is the subject of § 7 ; its reduced six-dimensional consequence is discussed in § 8 .
String dependence on a fixed ten-manifold
The calibrated value on Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) concerns a specified String datum. To see how much that specification matters, we now hold the sphere bundle fixed and vary its integral background data. The scalar can change while the metric, gauge connections and outward fibre period remain unchanged. This dependence is already contained in the relative integer of § 5 ; it does not require a different ten-dimensional geometry. We explain the change law and then two examples of Li and Sun [7 section 7] . The first lies in the Crowley–Goette domain, whereas the second isolates the additional information carried by the String datum outside that domain.
The pairing that detects a change of trivialization
Fix a closed spin seven-manifold B B B and its actual quaternionic line V V V , with underlying complex rank-two bundle U U U and c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) . We also write c c c for its Chern–Weil form at the fixed connection. Keep the Spin structure, metric, gauge connection, and all the characteristic and tautological morphisms of the disk construction fixed. Let h ˘ \breve{h} h ˘ be the integral differential trivialization of λ ˘ ( T B ) + c ˘ \breve{\lambda}(TB)+\breve{c} λ ˘ ( T B ) + c ˘ . Another such trivialization is obtained by adding a degree-three differential character α ˘ \breve{\alpha} α ˘ : the isomorphism classes of these trivializations form a torsor for H ^ 3 ( B ; Z ) \widehat{H}^{3}(B;\mathbb{Z}) H 3 ( B ; Z ) , as explained in Appendix § 13 . Write α ∈ H 3 ( B ; Z ) \alpha\in H^3(B;\mathbb{Z}) α ∈ H 3 ( B ; Z ) for its characteristic class and ω \omega ω for its curvature. Then
h ˘ ⟼ h ˘ + α ˘ , h ⟼ h + ω , d ω = 0. (86) \breve{h}\longmapsto\breve{h}+\breve{\alpha},\qquad h\longmapsto h+\omega,\qquad d\omega=0. \tag*{(86)} h ˘ ⟼ h ˘ + α ˘ , h ⟼ h + ω , d ω = 0. (86)
The equation d h = λ + c dh=\lambda+c d h = λ + c therefore remains the same. It does not by itself identify the two trivializations: the shift specifies their difference as integral differential data, including any part invisible to curvature.
The intrinsic expression of § 5 makes the effect especially transparent. Here ξ ( B , 4 U − 8 ) \xi(B,4U-8) ξ ( B , 4 U − 8 ) is the reduced Dirac invariant with the indicated virtual coefficient:
ϕ ( S ( V ) , s ˘ M ) = ξ ( B , 4 U − 8 ) − 1 3 ∫ B h ˘ ∧ c in R / Z . (87) \phi(S(V),\breve{s}_M)=\xi(B,4U-8)-\frac{1}{3}\int_B\breve{h}\wedge c\quad\text{in }\mathbb{R}/\mathbb{Z}. \tag*{(87)} ϕ ( S ( V ) , s ˘ M ) = ξ ( B , 4 U − 8 ) − 3 1 ∫ B h ˘ ∧ c in R / Z . (87)
The reduced Dirac invariant, including its kernel term, depends on the geometry and connection that we have held fixed. Only the second term changes. For the same spin/quaternionic filling W 8 W_8 W 8 , the bulk term in (64) is likewise fixed; the boundary subtraction changes by − ∫ B ω ∧ c -\int_B\omega\wedge c − ∫ B ω ∧ c . Since ω \omega ω represents the real image of α \alpha α , ∫ B ω ∧ c = ⟨ α ⌣ c , [ B ] ⟩ \int_B\omega\wedge c=\langle\alpha\smile c,[B]\rangle ∫ B ω ∧ c = ⟨ α ⌣ c , [ B ]⟩ . Thus
Δ K = − ⟨ α ⌣ c , [ B ] ⟩ , \Delta K=-\langle\alpha\smile c,[B]\rangle, Δ K = − ⟨ α ⌣ c , [ B ]⟩ ,
Δ ϕ = − 1 3 ⟨ α ⌣ c , [ B ] ⟩ in R / Z , (88) \Delta\phi=-\frac{1}{3}\langle\alpha\smile c,[B]\rangle\quad\text{in }\mathbb{R}/\mathbb{Z}, \tag*{(88)} Δ ϕ = − 3 1 ⟨ α ⌣ c , [ B ]⟩ in R / Z , (88)
O ( S ( V ) , s ˘ M ′ ) O ( S ( V ) , s ˘ M ) = exp ( − 2 π i 3 ⟨ α ⌣ c , [ B ] ⟩ ) . (89) \frac{\mathcal{O}(S(V),\breve{s}'_M)}{\mathcal{O}(S(V),\breve{s}_M)}
=\exp\left(-\frac{2\pi i}{3}\langle\alpha\smile c,[B]\rangle\right). \tag*{(89)} O ( S ( V ) , s ˘ M ) O ( S ( V ) , s ˘ M ′ ) = exp ( − 3 2 π i ⟨ α ⌣ c , [ B ]⟩ ) . (89)
The sign also follows directly from the relative cochain convention: the change of the boundary component of z z z is represented by ( 0 , α ) (0,\alpha) ( 0 , α ) , and relative integration subtracts the boundary integral. Both expressions therefore measure the same change of the input rather than a change of filling.
The disk cochain changes by π ∗ α ~ \pi^*\tilde{\alpha} π ∗ α ~ , while the tautological boundary morphism stays fixed. Consequently the induced differential String datum on M = S ( V ) M=S(V) M = S ( V ) changes by the same pullback, and its underlying integral class obeys
s ~ M ′ − s ~ M = π ∗ α ~ , [ s M ′ ] − [ s M ] = π ∗ α . (90) \tilde{s}'_M-\tilde{s}_M=\pi^*\tilde{\alpha}, \qquad[s'_M]-[s_M]=\pi^*\alpha. \tag*{(90)} s ~ M ′ − s ~ M = π ∗ α ~ , [ s M ′ ] − [ s M ] = π ∗ α . (90)
An integral String shift is therefore not generally a gauge transformation or an isomorphism of the original differential object. Equation (89) specifies which part of that shift this scalar detects.
It also specifies the information the scalar misses. A flat α ~ \tilde{\alpha} α ~ has zero curvature, so it leaves (87) unchanged. If α \alpha α is torsion, its top-dimensional pairing with c c c is zero: a torsion product cannot evaluate nontrivially on the integral fundamental class of the closed oriented seven-manifold. These are statements about response insensitivity, not equality of differential String objects. More generally, even a nonzero integral pairing leaves the exponentiated response unchanged when it is divisible by three. An equality of scalar values is thus a substantially weaker statement than an identification of the background data.
On the exact domain H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 H^3(B;\mathbb{Q})=H^4(B;\mathbb{Q})=0 H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 used in section 5, every integral degree-three class has zero real image and the pairing vanishes. This agrees with ϕ = 8 t B ( V ) \phi=8t_B(V) ϕ = 8 t B ( V ) : the scalar there is read from the quaternionic-line invariant, although the integral differential data remain part of its construction. The larger-domain formula (87) is what permits the following comparison between bases with and without those vanishings.
An evaluation on the unit tangent bundle of the four-sphere
Let x x x generate H 4 ( S 4 ; Z ) H^4(S^4;\mathbb{Z}) H 4 ( S 4 ; Z ) with ∫ S 4 x = 1 \int_{S^4}x=1 ∫ S 4 x = 1 , and take W 8 = D ( T S 4 ) W_8=D(TS^4) W 8 = D ( T S 4 ) with boundary the unit tangent sphere bundle B = S T S 4 B=STS^4 B = S T S 4 . Orient its tangent disk fibres positively, so e ( T S 4 ) = 2 x e(TS^4)=2x e ( T S 4 ) = 2 x . The Gysin sequence relates the cohomology of a sphere bundle to pullback, fibre integration and cup product with its Euler class. Its relevant portion here is
0 ⟶ H 3 ( B ; Z ) ⟶ Z → × 2 Z ⟶ H 4 ( B ; Z ) ⟶ 0. (91) 0\longrightarrow H^3(B;\mathbb{Z})\longrightarrow\mathbb{Z}\xrightarrow{\times2}\mathbb{Z}\longrightarrow H^4(B;\mathbb{Z})\longrightarrow0 . \tag*{(91)} 0 ⟶ H 3 ( B ; Z ) ⟶ Z × 2 Z ⟶ H 4 ( B ; Z ) ⟶ 0. (91)
It gives H 3 ( B ; Z ) = 0 H^3(B;\mathbb{Z})=0 H 3 ( B ; Z ) = 0 and H 4 ( B ; Z ) = Z / 2 H^4(B;\mathbb{Z})=\mathbb{Z}/2 H 4 ( B ; Z ) = Z /2 , hence precisely the two rational vanishings needed for the t t t comparison. The integral torsion in degree four has not been discarded.
Choose an actual quaternionic line on S 4 S^4 S 4 with c 2 = 2 x c_2=2x c 2 = 2 x and pull it back to W 8 W_8 W 8 . We write x x x also for its pullback. The tangent bundle of W 8 W_8 W 8 splits as the pullback of T S 4 ⊕ T S 4 TS^4\oplus TS^4 T S 4 ⊕ T S 4 ; since T S 4 ⊕ R TS^4\oplus\mathbb{R} T S 4 ⊕ R is trivial, λ ( T W 8 ) = 0 \lambda(TW_8)=0 λ ( T W 8 ) = 0 . For the positive Thom class u ∈ H 4 ( W 8 , B ; Z ) u\in H^4(W_8,B;\mathbb{Z}) u ∈ H 4 ( W 8 , B ; Z ) , normalized to integrate to one on each disk fibre, the map j j j to absolute cohomology satisfies j u = 2 x ju=2x j u = 2 x . Hence z = u z=u z = u is the relative class lifting λ W + c W \lambda_W+c_W λ W + c W . The absolute class 2 x 2x 2 x restricts to zero integrally on B B B , furnishing the required trivialization; the absence of H 3 ( B ; Z ) H^3(B;\mathbb{Z}) H 3 ( B ; Z ) fixes its integral class. Thom integration then yields
K = ⟨ 2 x ⌣ u , [ W 8 , B ] ⟩ = 2 , t B ( V ) = 1 12 , ϕ ( S ( V ) , s ~ M ) = 2 3 . (92) \mathsf{K}=\langle2x\smile u,[W_8,B]\rangle=2,\qquad t_B(V)=\frac{1}{12},\qquad\phi(S(V),\tilde{s}_M)=\frac{2}{3}. \tag*{(92)} K = ⟨ 2 x ⌣ u , [ W 8 , B ]⟩ = 2 , t B ( V ) = 12 1 , ϕ ( S ( V ) , s ~ M ) = 3 2 . (92)
The last two values are understood modulo Z \mathbb{Z} Z . This calculation uses an actual line and an integral boundary nullhomotopy, so both the relative formula and the Crowley–Goette comparison apply. It provides a second evaluation of the response, with the orientation accounted for by the positive tangent Thom class.
Three response values with the geometry held fixed
For a base with a detectable degree-three class, let Q − 1 Q_{-1} Q − 1 be an actual quaternionic line on S 4 S^4 S 4 with c 2 ( Q − 1 ) = − x c_2(Q_{-1})=-x c 2 ( Q − 1 ) = − x , and set
F 0 = R ⊕ Ad R ( Q − 1 ) , W 8 = D ( F 0 ) , B = S ( F 0 ) . (93) F_0=\mathbb{R}\oplus\operatorname{Ad}_{\mathbb{R}}(Q_{-1}),\qquad W_8=D(F_0),\qquad B=S(F_0). \tag*{(93)} F 0 = R ⊕ Ad R ( Q − 1 ) , W 8 = D ( F 0 ) , B = S ( F 0 ) . (93)
Here Ad R ( Q − 1 ) \operatorname{Ad}_{\mathbb{R}}(Q_{-1}) Ad R ( Q − 1 ) is the real rank-three adjoint bundle. Use ρ \rho ρ for the projection to S 4 S^4 S 4 and its extension over the disk bundle. The adjoint action lifts to Spin ( 3 ) \operatorname{Spin}(3) Spin ( 3 ) , making F 0 F_0 F 0 an oriented Spin four-plane bundle. Its real-line section forces e ( F 0 ) = 0 e(F_0)=0 e ( F 0 ) = 0 . The adjoint characteristic identity p 1 ( Ad R Q ) = − 4 c 2 ( Q ) p_1(\operatorname{Ad}_{\mathbb{R}}Q)=-4c_2(Q) p 1 ( Ad R Q ) = − 4 c 2 ( Q ) gives λ ( F 0 ) = 2 x \lambda(F_0)=2x λ ( F 0 ) = 2 x . Choose a second actual quaternionic line Q − 2 Q_{-2} Q − 2 with c 2 ( Q − 2 ) = − 2 x c_2(Q_{-2})=-2x c 2 ( Q − 2 ) = − 2 x and put V ~ = ρ ∗ Q − 2 \widetilde{V}=\rho^*Q_{-2} V = ρ ∗ Q − 2 on W 8 W_8 W 8 , with boundary restriction V = V ~ ∣ B V=\widetilde{V}|_B V = V ∣ B . The disk tangent splitting is the pullback of T S 4 ⊕ F 0 TS^4\oplus F_0 T S 4 ⊕ F 0 . The T S 4 TS^4 T S 4 summand has zero λ \lambda λ , so it gives
λ ( T W 8 ) = 2 ρ ∗ x , c W = − 2 ρ ∗ x , λ ( T W 8 ) + c W = 0 integrally . (94) \lambda(TW_8)=2\rho^*x,\qquad c_W=-2\rho^*x,\qquad\lambda(TW_8)+c_W=0\quad\text{integrally}. \tag*{(94)} λ ( T W 8 ) = 2 ρ ∗ x , c W = − 2 ρ ∗ x , λ ( T W 8 ) + c W = 0 integrally . (94)
The integral vanishing permits a geometric trivialization on W 8 W_8 W 8 ; choose one and restrict it to a boundary datum h ˘ 0 \breve{h}_0 h ˘ 0 . Its form component need not vanish, because the chosen characteristic forms need not vanish pointwise. What matters for the relative pairing is that this particular trivialization extends over the filling: its relative class is zero, so K 0 = 0 K_0=0 K 0 = 0 .
The first Gysin sequence, for ρ : B → S 4 \rho:B\to S^4 ρ : B → S 4 , now reads
0 ⟶ H 3 ( B ; Z ) → ρ ∗ Z → ⌣ e ( F 0 ) = 0 Z → ρ ∗ H 4 ( B ; Z ) ⟶ 0. (95) 0\longrightarrow H^3(B;\mathbb{Z})\xrightarrow{\rho_*}\mathbb{Z}\xrightarrow{\smile e(F_0)=0}\mathbb{Z}\xrightarrow{\rho^*}H^4(B;\mathbb{Z})\longrightarrow0. \tag*{(95)} 0 ⟶ H 3 ( B ; Z ) ρ ∗ Z ⌣ e ( F 0 ) = 0 Z ρ ∗ H 4 ( B ; Z ) ⟶ 0. (95)
Thus H 3 ( B ; Z ) ≅ Z H^3(B;\mathbb{Z})\cong\mathbb{Z} H 3 ( B ; Z ) ≅ Z and H 4 ( B ; Z ) ≅ Z H^4(B;\mathbb{Z})\cong\mathbb{Z} H 4 ( B ; Z ) ≅ Z . The map ρ ∗ \rho_* ρ ∗ integrates over the oriented sphere fibre and is an isomorphism here. Choose its primitive generator α \alpha α so that ρ ∗ α = 1 \rho_*\alpha=1 ρ ∗ α = 1 ; the pullback ρ ∗ x \rho^*x ρ ∗ x is likewise a primitive generator in degree four. The projection formula then computes the pairing needed by the change law:
⟨ α ⌣ c , [ B ] ⟩ = − 2 ⟨ ( ρ ∗ α ) ⌣ x , [ S 4 ] ⟩ = − 2. (96) \langle\alpha\smile c,[B]\rangle=-2\langle(\rho_*\alpha)\smile x,[S^4]\rangle=-2. \tag*{(96)} ⟨ α ⌣ c , [ B ]⟩ = − 2 ⟨( ρ ∗ α ) ⌣ x , [ S 4 ]⟩ = − 2. (96)
Choose a differential refinement α ˘ \breve{\alpha} α ˘ and use h ˘ n = h ˘ 0 + n α ˘ \breve{h}_n=\breve{h}_0+n\breve{\alpha} h ˘ n = h ˘ 0 + n α ˘ . Equations (88) and (96) give
K n = 2 n , ϕ n = 2 n 3 ( m o d Z ) , O ( M , s ˘ n ) = exp ( 4 π i n / 3 ) . (97) K_n=2n,\qquad\phi_n=\frac{2n}{3}\pmod{\mathbb{Z}},\qquad\mathcal{O}(M,\breve{s}_n)=\exp(4\pi in/3). \tag*{(97)} K n = 2 n , ϕ n = 3 2 n ( mod Z ) , O ( M , s ˘ n ) = exp ( 4 π in /3 ) . (97)
For n = 0 , 1 , 2 n=0,1,2 n = 0 , 1 , 2 these are the three distinct values 1 1 1 , e 4 π i / 3 e^{4\pi i/3} e 4 π i /3 , e 2 π i / 3 e^{2\pi i/3} e 2 π i /3 .
The phase calculation alone does not prove that the induced integral String classes on M M M differ: a pullback can kill a base class. This is the role of the second Gysin sequence, for π : M = S ( V ) → B \pi:M=S(V)\to B π : M = S ( V ) → B . Our quaternionic orientation gives e ( V R ) = − c = 2 ρ ∗ x e(V_{\mathbb{R}})=-c=2\rho^*x e ( V R ) = − c = 2 ρ ∗ x , so its relevant terms are
0 ⟶ H 3 ( B ; Z ) → π ∗ H 3 ( M ; Z ) ⟶ H 0 ( B ; Z ) → ⌣ 2 ρ ∗ x H 4 ( B ; Z ) . (98) 0\longrightarrow H^3(B;\mathbb{Z})\xrightarrow{\pi^*}H^3(M;\mathbb{Z})\longrightarrow H^0(B;\mathbb{Z})\xrightarrow{\smile2\rho^*x}H^4(B;\mathbb{Z}). \tag*{(98)} 0 ⟶ H 3 ( B ; Z ) π ∗ H 3 ( M ; Z ) ⟶ H 0 ( B ; Z ) ⌣ 2 ρ ∗ x H 4 ( B ; Z ) . (98)
The final map is multiplication by two between copies of Z \mathbb{Z} Z , and is injective. Exactness therefore makes π ∗ \pi^* π ∗ an isomorphism in degree three. Equation (90) now shows that all the classes s n s_n s n and s m s_m s m differ by the nonzero class ( n − m ) π ∗ α (n-m)\pi^*\alpha ( n − m ) π ∗ α whenever n ≠ m n \ne m n = m . In particular, s n s_n s n and s n + 3 s_{n+3} s n + 3 remain distinct integral classes even though their scalar values agree. The response detects the integer parameter only modulo three; the integral String class retains the whole parameter.
Throughout this family the ten-manifold M M M , its sphere projection, Spin data, metric, gauge bundles, connections and tautological frames are held fixed. The changed data are the base integral differential trivialization and the String datum it induces on M M M . The change of H ˇ N \check{H}_N H ˇ N is pulled back from B B B and has zero restriction to every disk fibre. All backgrounds consequently retain the outward fibre period − 2 -2 − 2 of section 3, at the specified connection. There is no use here of the special right-framed + 1 +1 + 1 transport on Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) .
Both H 3 ( B ; Q ) H^3(B;\mathbb{Q}) H 3 ( B ; Q ) and H 4 ( B ; Q ) H^4(B;\mathbb{Q}) H 4 ( B ; Q ) are nonzero in this example. Equation (97) uses the general relative scalar formula, not three different values of t B ( V ) t_B(V) t B ( V ) in its original domain. The three values are responses to distinct background data; they do not count vacua or establish a phase transition. What the calculation establishes is that the underlying ten-manifold and a bare fibre period do not together determine this response. The degree-three pairing makes that missing information explicit, and leads to the six-dimensional anomaly interpretation of the same corrected base expression in section 8.
Six-dimensional response and integral lattice factorization
The intrinsic expression in Section 5 gives the ten-dimensional scalar a second interpretation: it is a closed seven-dimensional anomaly response for a virtual chiral system in six dimensions. This interpretation uses the same quaternionic line and the same integral differential trivialization that entered the sphere-bundle calculation. In particular, the sensitivity to the String datum found in Section 7 remains part of the background on which the reduced response is evaluated.
Let U U U be the complex rank-two bundle underlying the actual quaternionic line V V V , and retain c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) and λ = p 1 ( T B ) / 2 \lambda=p_1(TB)/2 λ = p 1 ( T B ) /2 . Define
R + = 4 U − 8 , I + = 1 3 c ( λ + c ) , (99) R_+ = 4U - 8,\qquad I_+ = \frac{1}{3}c(\lambda+c), \tag*{(99)} R + = 4 U − 8 , I + = 3 1 c ( λ + c ) , (99)
A 7 ( B , V , h ˇ ) = exp ( 2 π i [ ξ ( B , R + ) − 1 3 ∫ B h ˇ ∧ c ] ) . (100) A_7(B,V,\check{h}) = \exp\left(2\pi i\left[\xi(B,R_+) - \frac{1}{3}\int_B \check{h}\wedge c\right]\right). \tag*{(100)} A 7 ( B , V , h ˇ ) = exp ( 2 π i [ ξ ( B , R + ) − 3 1 ∫ B h ˇ ∧ c ] ) . (100)
Here the integer 8 denotes eight trivial complex summands. Without an additional reality condition, R + R_+ R + describes four complex Weyl doublets of one chirality and eight complex Weyl singlets of the opposite chirality. Subtraction records their opposite contributions to the anomaly. It is a virtual anomaly system, rather than a specified interacting matter theory. The quaternionic origin of U U U does not impose an additional reality condition on these six-dimensional Weyl fields.
The reduced invariant in (100) is the complex invariant ξ = ( η A P S + dim C ker D ) / 2 \xi=(\eta_{\mathrm{APS}}+\dim_{\mathbb{C}}\ker D)/2 ξ = ( η APS + dim C ker D ) /2 , evaluated additively on the virtual coefficient. Its kernel term is retained. There is no additional real Pfaffian half multiplying this seven-dimensional expression. The real normalization of the original eleven-dimensional operator has already entered the direct APS reduction. The eight-form I ^ + \widehat{I}_{+} I + is consequently the index density appropriate to this complex coefficient, with the chirality sign fixed by (100) .
One can see its normalization directly. Through degree eight, the rank-two SU(2) bundle has
ch ( U ) = 2 − c + c 2 12 + ⋯ , A ^ 4 = − λ 12 . \operatorname{ch}(U) = 2-c+\frac{c^{2}}{12}+\cdots,\qquad\widehat{A}_{4}=-\frac{\lambda}{12}. ch ( U ) = 2 − c + 12 c 2 + ⋯ , A 4 = − 12 λ .
The rank of 4 U − 8 4U-8 4 U − 8 vanishes, so the purely gravitational degree-eight term proportional to that rank cancels. Its remaining degree-four and degree-eight Chern character terms are − 4 c -4c − 4 c and c 2 / 3 c^{2}/3 c 2 /3 . Thus
[ A ^ ch ( R + ) ] 8 = ( − λ 12 ) ( − 4 c ) + c 2 3 = 1 3 c ( λ + c ) . (101) [\widehat{A}\operatorname{ch}(R_{+})]_{8}=\left(-\frac{\lambda}{12}\right)(-4c)+\frac{c^{2}}{3}=\frac{1}{3}c(\lambda+c). \tag*{(101)} [ A ch ( R + ) ] 8 = ( − 12 λ ) ( − 4 c ) + 3 c 2 = 3 1 c ( λ + c ) . (101)
This calculation fixes both the mixed gravitational–gauge coefficient and the pure gauge coefficient; agreement of only the latter would not specify the response.
The local correction uses the geometric component h h h of the full datum
d h ˇ = λ ˇ + c ˇ , d h = λ + c . (102) d\check{h}=\check{\lambda}+\check{c},\qquad dh=\lambda+c. \tag*{(102)} d h ˇ = λ ˇ + c ˇ , d h = λ + c . (102)
At the form level, d ( h ∧ c ) = c ( λ + c ) d(h\wedge c)=c(\lambda+c) d ( h ∧ c ) = c ( λ + c ) because d c = 0 dc=0 d c = 0 . The subtraction in (100) therefore cancels the local variation of the eta response on these backgrounds. The full differential equation also fixes the integral nullhomotopy and its compatibility with the connection data; a freely chosen real primitive would not supply them. Local cancellation leaves a closed response whose value can still depend on that integral datum. The eight-dimensional spin/gauge filling used to evaluate it extends the actual line with the prescribed boundary identification, without requiring h ˇ \check{h} h ˇ to extend as a trivialization over the filling.
For these data the direct calculation and the relative pairing give
A 7 ( B , V , h ˇ ) = O ( S ( V ) , s ˇ M ) = exp ( 2 π i K 3 ) . (103) \mathcal{A}_{7}(B,V,\check{h})=\mathcal{O}(S(V),\check{s}_{M})=\exp\left(2\pi i\frac{K}{3}\right). \tag*{(103)} A 7 ( B , V , h ˇ ) = O ( S ( V ) , s ˇ M ) = exp ( 2 π i 3 K ) . (103)
The sphere-bundle formula [7, Theorem 1.1; Sections 4–5 and 8], derived in § 5 , identifies the closed response on the permitted backgrounds, with their orientation and differential data. It is stronger than matching local polynomials, since the actual filling evaluates the eta term, but does not assert an equivalence of six- and ten-dimensional theories. No rational H 3 H^{3} H 3 or H 4 H^{4} H 4 vanishing is needed here. Those restrictions belong to the comparison with the Crowley–Goette invariant, rather than to (103) .
The cubic character and the half-doublet normalization
Integrality of K K K immediately implies A 7 3 = 1 \mathcal{A}_{7}^{3}=1 A 7 3 = 1 on every permitted background. To decide whether a smaller number of copies has trivial response everywhere, one must also know that a primitive third root occurs. Section 6 provides that test: the actual quaternionic line in the Sp ( 2 ) → S 7 \operatorname{Sp}(2)\to S^{7} Sp ( 2 ) → S 7 fibration, with the specified orientation and transported datum, has K = − 2 K=-2 K = − 2 . Hence its response is exp ( 2 π i / 3 ) \exp(2\pi i/3) exp ( 2 π i /3 ) . The integrality argument and this example together prove, for a nonnegative integer r r r , that the stacked response is identically one on the permitted background category precisely when three divides r r r . Writing A 7 r ≡ 1 A_7^r \equiv1 A 7 r ≡ 1 for this universal statement, we have
A 7 r ≡ 1 ⟺ 3 ∣ r . (104) A_7^r \equiv1 \Longleftrightarrow3 \mid r. \tag*{(104)} A 7 r ≡ 1 ⟺ 3 ∣ r . (104)
For the forward implication, the response of r r r copies on the primitive background is one only when 3 ∣ r 3 \mid r 3 ∣ r . For the reverse implication, r = 3 k r = 3k r = 3 k makes the exponent k K kK k K integral on every background. Thus the sufficiency is universal within the stated domain, while the necessity is witnessed by the calibrated example. Stacking copies means multiplying these closed anomaly responses; this calculation does not classify full quantum field theories or their states.
The coefficient has a known six-dimensional normalization precedent. Lee–Tachikawa [10, §2.2, (2.14); §3.3, (3.14)–(3.17)] give the gauge-dependent part of the primitive half-doublet density as
I 8 h a l f = 1 24 c ( λ + c ) , I 8 + = 8 I 8 h a l f . (105) I_8^{\mathrm{half}} = \frac{1}{24}c(\lambda+c), \qquad I_8^+ = 8I_8^{\mathrm{half}}. \tag*{(105)} I 8 half = 24 1 c ( λ + c ) , I 8 + = 8 I 8 half . (105)
The half-doublet includes a reality condition; our virtual complex coefficient has no such condition. Its eight neutral singlet subtractions cancel the purely gravitational term, so the comparison in (105) concerns the remaining gauge-dependent density. Their modified anomaly response detects a generator with value exp ( 2 π i / 12 ) \exp(2\pi i/12) exp ( 2 π i /12 ) , giving an order-twelve condition for copies of that primitive system. Eight copies per R + R_+ R + system give the same arithmetic condition as (104) :
12 ∣ 8 r ⟺ 3 ∣ r . (106) 12 \mid8r \Longleftrightarrow3 \mid r. \tag*{(106)} 12 ∣ 8 r ⟺ 3 ∣ r . (106)
The divisibility in (104) is thus a specialization of the known six-dimensional normalization condition.
The backgrounds in the two calculations differ. Lee–Tachikawa add a three-form satisfying d H = c \mathrm{d}H = c d H = c ; their local correction is − ∫ H ∧ ( c + λ ) / 24 -\int H \wedge(c+\lambda)/24 − ∫ H ∧ ( c + λ ) /24 . In this companion the prescribed integral differential datum trivializes λ + c \lambda+c λ + c , and the local correction is − ∫ h ∧ c / 3 -\int h \wedge c/3 − ∫ h ∧ c /3 . Their seven-form integrands have proportional exterior derivatives, but the objects trivialized and their global data are different. The density comparison therefore explains the coefficient and the resulting divisibility without identifying the two anomaly theories on all backgrounds or families. In particular, the primitive Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) evaluation used above is the one already computed with this companion’s background datum.
This distinction also keeps local cancellation separate from a dynamical tensor construction. Lee–Tachikawa’s discussion following (3.17) explains why cancellation for twelve copies with their classical three-form does not alone define the path integral over gauge and two-form fields. The closed character is one part of the anomaly problem. Its triviality does not supply the missing quantum tensor data.
Universal coefficient matching in an integral lattice
In six-dimensional theories with chiral two-form fields, the lattice records charges of strings coupled to those fields. Locality of probe-string surface operators gives an integral pairing, and consistency under dimensional reduction requires a self-dual, or unimodular, charge lattice [19 §2.3, (2.9)–(2.11)] . Seiberg–Taylor’s even signature-( 1 , 1 ) (1, 1) ( 1 , 1 ) example has the hyperbolic pairing used below [19 §2.4, (2.16)] . This motivates the arithmetic setting for our coefficient-matching question; it does not introduce tensor fields into the original background.
The same divisibility has a precise arithmetic formulation in the usual tensor Green–Schwarz normalization. Monnier–Moore write the eight-form as 1 2 ( Y , Y ) \frac{1}{2}(Y,Y) 2 1 ( Y , Y ) and, for a single simple gauge factor, use Y = a p 1 / 4 − b c Y=a\,p_1/4-bc Y = a p 1 /4 − b c [11 (2.5), (2.10)–(2.12)] . Here Y Y Y is a lattice-valued degree-four polynomial, distinct from the degree-eight Y 8 Y_8 Y 8 of the ten-dimensional calculation. The vectors a a a and b b b are respectively the gravitational and gauge coefficients in this normalization. Monnier–Moore–Park relate their integrality to string-charge quantization and distinguish conditions for the covering gauge group from those for the full gauge global form [12 §2.3, constraints 9, 9* and 10] . Those integrality constraints use a generalized completeness hypothesis: the theory is assumed to admit arbitrary spin spacetimes with arbitrary smooth gauge-field configurations. These background assumptions enter their physical derivation of charge quantization. In that normalization, we ask whether r I 8 + rI_8^+ r I 8 + admits an integral lattice factorization. The scalar calculation does not establish the broader background assumptions used in the supergravity analysis.
Take c c c to be the primitive integral second Chern class of an actual S U ( 2 ) = S p ( 1 ) \mathrm{SU}(2)=\mathrm{Sp}(1) SU ( 2 ) = Sp ( 1 ) bundle. An integral lattice Λ \Lambda Λ means a free abelian group with an integer-valued symmetric nondegenerate pairing. Require a , b ∈ Λ a,b\in\Lambda a , b ∈ Λ and require a a a to be characteristic:
( a , v ) ≡ ( v , v ) ( m o d 2 ) for every v ∈ Λ . (107) (a,v)\equiv(v,v)\pmod{2}\qquad\text{for every }v\in\Lambda. \tag*{(107)} ( a , v ) ≡ ( v , v ) ( mod 2 ) for every v ∈ Λ. (107)
The desired identity is
1 2 ( Y , Y ) = r I 8 + , Y = a 2 λ − b c . (108) \frac{1}{2}(Y,Y)=rI_8^+,\qquad Y=\frac{a}{2}\lambda-bc. \tag*{(108)} 2 1 ( Y , Y ) = r I 8 + , Y = 2 a λ − b c . (108)
It is an identity of universal polynomials in λ \lambda λ and c c c . One must compare its coefficients before restricting to a background on which λ + c \lambda+c λ + c is trivialized. Substituting that relation first would erase the information the factorization is meant to test.
Expanding the left-hand side gives
1 2 ( Y , Y ) = ( a , a ) 8 λ 2 − ( a , b ) 2 λ c + ( b , b ) 2 c 2 . (109) \frac{1}{2}(Y,Y)=\frac{(a,a)}{8}\lambda^2-\frac{(a,b)}{2}\lambda c+\frac{(b,b)}{2}c^2. \tag*{(109)} 2 1 ( Y , Y ) = 8 ( a , a ) λ 2 − 2 ( a , b ) λ c + 2 ( b , b ) c 2 . (109)
Comparison with r c ( λ + c ) / 3 rc(\lambda+c)/3 r c ( λ + c ) /3 therefore requires
( a , a ) = 0 , ( a , b ) = − 2 r 3 , ( b , b ) = 2 r 3 . (110) (a,a)=0,\qquad(a,b)=-\frac{2r}{3},\qquad(b,b)=\frac{2r}{3}. \tag*{(110)} ( a , a ) = 0 , ( a , b ) = − 3 2 r , ( b , b ) = 3 2 r . (110)
The last equation already proves necessity: ( b , b ) (b,b) ( b , b ) is an integer, so 3 ∣ r 3\mid r 3 ∣ r . This implication uses neither unimodularity nor the characteristic condition. Those conditions matter for constructing the converse with the specified gravitational coefficient.
For r = 3 k r=3k r = 3 k , take the rank-two hyperbolic lattice with basis e , f e,f e , f and pairing
( e , e ) = ( f , f ) = 0 , ( e , f ) = 1 , a = 2 k f , b = − e − k f . (111) (e,e)=(f,f)=0,\qquad(e,f)=1,\qquad a=2kf,\qquad b=-e-kf. \tag*{(111)} ( e , e ) = ( f , f ) = 0 , ( e , f ) = 1 , a = 2 k f , b = − e − k f . (111)
Its Gram matrix has determinant − 1 -1 − 1 , so the lattice is unimodular. For v = m e + n f v = me + nf v = m e + n f , ( v , v ) = 2 m n (v,v) = 2mn ( v , v ) = 2 mn is even, so it is also even. Both coefficients are integral, and
( a , v ) − ( v , v ) = 2 m ( k − n ) ∈ 2 Z . (112) (a,v) - (v,v) = 2m(k-n) \in2\mathbb{Z}. \tag*{(112)} ( a , v ) − ( v , v ) = 2 m ( k − n ) ∈ 2 Z . (112)
Thus a a a is characteristic. With these choices the degree-four polynomial becomes
Y = e c + k f ( λ + c ) , 1 2 ( Y , Y ) = k c ( λ + c ) = r I 8 + . (113) Y = ec + kf(\lambda+ c), \qquad\frac{1}{2}(Y,Y) = kc(\lambda+ c) = rI_8^+. \tag*{(113)} Y = ec + k f ( λ + c ) , 2 1 ( Y , Y ) = k c ( λ + c ) = r I 8 + . (113)
Only the cross-pairing contributes, which makes both the sign of b b b and the absence of a λ 2 \lambda^2 λ 2 term transparent. This construction also covers r = 0 r = 0 r = 0 with k = 0 k = 0 k = 0 . Consequently, the universal factorization with integral gauge coefficient and characteristic gravitational coefficient exists precisely for multiples of three, and an even unimodular hyperbolic lattice suffices [7 Section 8, Proposition 8.1] .
The conclusion concerns this polynomial and these integral coefficients. The primitive class and actual gauge global form are essential to its meaning; a change of matter content, gauge group or anomaly sector changes the question. Moreover, the tensor formulation uses a lattice-valued differential trivialization d H ˇ = Y ˇ d\check{H} = \check{Y} d H ˇ = Y ˇ , whose curvature equation is d H = Y dH = Y d H = Y . Monnier–Moore’s construction shifts Y ˇ \check{Y} Y ˇ by 1 2 ν ⊗ a \frac{1}{2}\nu\otimes a 2 1 ν ⊗ a , where ν \nu ν represents the degree-four Wu class. On spin manifolds this class equals the fourth Stiefel–Whitney class w 4 w_4 w 4 . The cochain H ˇ \check{H} H ˇ has a compatible shift [11 §3.2, (3.9)–(3.12)] . These data differ from the integral trivialization used by our original scalar in (102) . The explicit lattice identity does not construct quantum self-dual tensors or establish the remaining matter and gravitational conditions of a complete supergravity theory. It supplies an arithmetic consequence of the reduced response, with the known tensor normalization kept visible.
The identification with a particular published heterotic prescription requires a different comparison: the actual real operator, its line and its differential transport must agree. § 9 addresses those inputs. The lattice calculation alone supplies none of them.
Comparison with the heterotic fermion prescription
The response computed above can be compared with the ordered heterotic Green–Schwarz phase of Tachikawa–Zhang [3 ] . The relevant object is the Spin ( 32 ) / Z 2 \operatorname{Spin}(32)/\mathbb{Z}_2 Spin ( 32 ) / Z 2 phase divided by the E 8 × E 8 E_8 \times E_8 E 8 × E 8 phase, evaluated on the same boundary geometry and differential String datum. Identifying this object requires more than recovering its known third-root value: the closed fermion systems, their chirality, their lines and their reference normalization must agree. The comparison holds on the full filling data, before either the local integral or closed eta invariant is specialized [7 §9] . Its operator and connection content is separate from the conditional integral topological modular forms (TMF) comparison in § 10 .
The ordered phase and its reference
The physical input is the fermion prescription used in the published analysis of Tachikawa–Zhang [3 §2.1, §§4.2–4.4] . The two supersymmetric theories have adjoint gauginos of the same chirality and identical gravitational fermion and ghost sectors. Their anomaly is the real fermion eta anomaly of these systems, with no additional independent invertible sector. For each theory the differential H 3 H_3 H 3 -dependent section obeys the consistency law of [3 equation (67)] , takes the empty background to the unit, and is paired with that same theory’s trivial-gauge spin reference as in [3 equation (70)] . The Hermitian pairing is conjugate-linear in its first slot. All of these clauses specify the observable; they are not an assertion that every interacting theory is determined by its massless spectrum. An independently added flat anomaly sector would change this prescription while leaving its local polynomial unchanged.
For a bordism A : M 0 → M 1 A:M_0 \to M_1 A : M 0 → M 1 , the section law for either theory has the form
v j ( M 1 ) = exp ( 2 π i ∫ A H 3 Y 8 , j ) Z j ( A ) v j ( M 0 ) , v j ( ∅ ) = 1 , v_j(M_1)=\exp\left(2\pi i\int_A H_3Y_{8,j}\right)Z_j(A)v_j(M_0),\qquad v_j(\varnothing)=1, v j ( M 1 ) = exp ( 2 π i ∫ A H 3 Y 8 , j ) Z j ( A ) v j ( M 0 ) , v j ( ∅ ) = 1 ,
where j j j denotes that theory and Z j ( A ) Z_j(A) Z j ( A ) is its anomaly evolution map. The law relates vectors in anomaly lines, rather than assigning an absolute complex number to the fermion integral. The spin reference supplies the comparison vector in the same line. Applied to a filling from the empty manifold, the law fixes the normalized section; its first-slot pairing then conjugates the local factor. Taking the ratio of the two theories uses their own sections and references, and the common X 4 X_4 X 4 turns the difference of Y 8 , j Y_{8,j} Y 8 , j into Y 8 Y_8 Y 8 . This explains why the reference and the order of the quotient are essential to the sign of the scalar.
Let M M M be a connected closed spin ten-manifold with framed trivial gauge connections, differential String data and specified tangent geometry. Use a product-collar differential H \mathcal{H} H -filling N N N and a trivial-gauge spin reference W r e f W_{\mathrm{ref}} W ref with their specified boundary Spin and differential identifications. Tachikawa–Zhang’s consistency law and reference pairing, followed by its prescribed gluing, give [3 equations (71)–(73)]
O T Z ( M ) = exp ( − 2 π i ∫ N H 3 Y 8 ) I d i f f ( C ) , C = N ‾ ∪ M W r e f . (114) \mathcal{O}_{\mathrm{TZ}}(M)=\exp\left(-2\pi i\int_N H_3Y_8\right)I_{\mathrm{diff}}(C),\qquad C=\overline{N}\cup_M W_{\mathrm{ref}}. \tag*{(114)} O TZ ( M ) = exp ( − 2 π i ∫ N H 3 Y 8 ) I diff ( C ) , C = N ∪ M W ref . (114)
Here I d i f f ( C ) I_{\mathrm{diff}}(C) I diff ( C ) is the quotient of the closed anomaly partition functions in the stated order. The negative local sign and the orientation reversal of N N N both come from the first-slot conjugation. On the common cover the two X 4 X_4 X 4 polynomials agree, so Y 8 Y_8 Y 8 is the full difference in (7) ; the reference does not replace this difference by a first-root specialization.
The reference W r e f W_{\mathrm{ref}} W ref is a spin filling with trivial gauges; it is not required to fill the differential H \mathcal{H} H structure. Thus C C C is a closed spin/gauge manifold, rather than necessarily a closed differential H \mathcal{H} H manifold to which closed cancellation would apply. A nontrivial scalar is possible even though the section on H \mathcal{H} H boundaries is independent of its H \mathcal{H} H filling. Replacing the reference by an H \mathcal{H} H -filling in this argument would change the comparison being made. The actual seam identifications also matter: trivial characteristic classes on the boundary alone do not specify the framed Spin connection or the differential morphism.
The real Pfaffian contraction in (114) is constructed in Appendix § 14 .§ 5 . It matches the Witten–Yonekura and Freed–Moore allowed boundary domains and spectral-cut Pfaffian frames, retaining kernels and the actual Spin and gauge seam maps. The transported state in (B.19) includes the finite stable tangent correction; its Hermitian contraction in (B.20) reverses the first filling.
The Dai–Freed erratum [20 ] corrects the odd-dimensional collar operator by a J − 1 J^{-1} J − 1 factor, with J J J the outward normal Clifford action, and amends orientation reversal and graded gluing. These complex determinant-line formulas do not replace the real component parities. The real construction retains ordered duality of its component lines and the specified seam identifications, rather than taking a square root of a complex gluing identity. It supplies the contraction for the stated fermion prescription; the closed-operator comparison below does not assert an equivalence of complete anomaly functors.
Parallel real branching and the fermion line
The gauge backgrounds used in the calculation lift to the actual common cover G = Spin ( 16 ) × Spin ( 16 ) G = \operatorname{Spin}(16) \times\operatorname{Spin}(16) G = Spin ( 16 ) × Spin ( 16 ) . With its induced connections, the two real adjoint bundles decompose as
ad ( Spin ( 32 ) ) ∣ G = Λ 2 E ⊕ ( E ⊗ R E ′ ) ⊕ Λ 2 E ′ , \operatorname{ad}(\operatorname{Spin}(32))\big|_{G} = \Lambda^{2}E \oplus(E \otimes_{\mathbb{R}} E') \oplus\Lambda^{2}E', ad ( Spin ( 32 )) G = Λ 2 E ⊕ ( E ⊗ R E ′ ) ⊕ Λ 2 E ′ ,
ad ( E 8 × E 8 ) ∣ G = Λ 2 E ⊕ S R + ( E ) ⊕ Λ 2 E ′ ⊕ S R + ( E ′ ) . (115) \operatorname{ad}(E_{8} \times E_{8})\big|_{G} = \Lambda^{2}E \oplus S^{+}_{\mathbb{R}}(E) \oplus\Lambda^{2}E' \oplus S^{+}_{\mathbb{R}}(E'). \tag*{(115)} ad ( E 8 × E 8 ) G = Λ 2 E ⊕ S R + ( E ) ⊕ Λ 2 E ′ ⊕ S R + ( E ′ ) . (115)
The first equality is the skew-matrix block decomposition. For the second, the 112 roots ± e i ± e j \pm e_i \pm e_j ± e i ± e j and eight Cartan directions give the 120-dimensional D 8 D_{8} D 8 adjoint; the remaining 128 roots 1 2 ( ± e 1 ± ⋯ ± e 8 ) \frac{1}{2}(\pm e_1 \pm\cdots\pm e_8) 2 1 ( ± e 1 ± ⋯ ± e 8 ) with even sign parity give the real half-spin module. The embedding fixes which half-spin module occurs. These are equivariant real-module isomorphisms. Invariant inner products make them orthogonal, and equivariance makes them parallel for every connection induced from the common cover.
Tensoring these maps with the spacetime spinors therefore intertwines the actual closed Dirac operators. It preserves their kernels and flat holonomy, as well as the characteristic forms. The common adjoint blocks cancel spectrally, and the identical gravitational fermion and ghost factors cancel against their inverses. The remaining coefficient is the real virtual bundle
Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , \Delta_{\mathbb{R}} = E \otimes_{\mathbb{R}} E' - S^{+}_{\mathbb{R}}(E) - S^{+}_{\mathbb{R}}(E'), Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) ,
up to the common gaugino chirality. This stronger statement is what allows a comparison of global phases: equality of Chern characters alone would not control the spectrum or its flat dependence.
The sign of the phase also depends on the line in which the fermion integral takes values. For a real twisting coefficient R R R , fix positive-index chirality by I ( R ) = 1 2 [ A ^ ( R C ) ] 12 I(R) = \frac{1}{2}[\widehat{A}(\mathbb{R}_{C})]_{12} I ( R ) = 2 1 [ A ( R C ) ] 12 and denote the associated boundary Pfaffian line by P R \mathcal{P}_{R} P R . Freed–Moore place their real half-eta state exp ( + π i ξ R ) \exp(+\pi i\xi_{R}) exp ( + π i ξ R ) in P R ∗ \mathcal{P}_{R}^{*} P R ∗ [16 equation (3.11)] ; the positive fermion integral itself belongs to P R \mathcal{P}_{R} P R [16 equations (4.33)–(4.34)] . Only this line assignment is used, rather than their later conjectural M-theory cancellation. Tachikawa–Zhang put the fermion integral and anomaly state in the same line. Their positive-index closed fermion phase is accordingly the dual state,
χ R ( C ) = exp ( − π i ξ R ( C ) ) , ξ R = η A P S ( D R ) + dim ker D R 2 . (116) \chi_{R}(C) = \exp(-\pi i\xi_{R}(C)), \qquad\xi_{R} = \frac{\eta_{\mathrm{APS}}(D_{R}) + \dim\ker D_{R}}{2}. \tag*{(116)} χ R ( C ) = exp ( − π i ξ R ( C )) , ξ R = 2 η APS ( D R ) + dim ker D R . (116)
Witten–Yonekura’s negative eta exponent describes inflow in the line dual to the boundary fermion [17 equations (2.10), (2.52)] . To compare it with (116) , the closed operator must be translated along with the line. With Freed–Moore’s skew Clifford action c ( v ) c(v) c ( v ) and normalized odd volume action ω \omega ω , the translation is γ W Y ( v ) = i ω c ( v ) \gamma_{\mathrm{WY}}(v) = i\omega c(v) γ WY ( v ) = iω c ( v ) . With outward unit normal ν \nu ν and boundary volume action ω ∂ \omega_{\partial} ω ∂ , the collar relation is γ W Y ( ν ) = − i ω ∂ \gamma_{\mathrm{WY}}(\nu) = -i\omega_{\partial} γ WY ( ν ) = − i ω ∂ , so the displayed boundary chiralities agree, but the closed odd operators have opposite signs:
D W Y = − d F M , η W Y = − 2 ξ R + 2 h R , D_{\mathrm{WY}} = -d_{\mathrm{FM}}, \qquad\eta_{\mathrm{WY}} = -2\xi_{R} + 2h_{R}, D WY = − d FM , η WY = − 2 ξ R + 2 h R ,
e − i π η W Y / 2 = e + π i ξ R , h R = dim C ker d F M ∈ 2 Z . (117) e^{-i\pi\eta_{\mathrm{WY}}/2} = e^{+\pi i\xi_{R}}, \qquad h_{R} = \dim_{\mathbb{C}}\ker d_{\mathrm{FM}} \in2\mathbb{Z}. \tag*{(117)} e − iπ η WY /2 = e + π i ξ R , h R = dim C ker d FM ∈ 2 Z . (117)
Here D D D is the Clifford-module Dirac operator and d F M = ω D ∣ M 0 d_{\mathrm{FM}} = \omega D|_{M_{0}} d FM = ω D ∣ M 0 is its odd self-adjoint operator on the even Clifford-module summand. The volume action commutes with the Clifford action. The induced boundary grading intertwines the oppositely signed tangential operators without reversing their graded index. The dictionary therefore preserves the boundary fermion while changing the sign of the closed odd operator, which is precisely the distinction needed for the eta comparison. The kernel multiplicity is even because closed real-twisted eleven-dimensional Dirac eigenspaces are quaternionic. Thus the inflow state agrees with Freed–Moore’s inverse-line state, whose dual gives (116) . The convention η = 2 ξ \eta= 2\xi η = 2 ξ discussed after [17 ] , equation (2.19) applies to Witten–Yonekura’s own odd operator; it cannot identify differently signed operators merely because their boundary chirality is written the same way.
Direct sums add reduced eta invariants including their kernels, and virtual subtraction inverts phases. Quaternionic spectral crossings introduce no jump in the real phase, whereas ξ ( − D ) = − ξ ( D ) + h R \xi(-D) = -\xi(D) + h_{R} ξ ( − D ) = − ξ ( D ) + h R gives its complex conjugate under operator reversal. The real-valued representatives are combined before reduction; no half of an arbitrary R / Z \mathbb{R}/\mathbb{Z} R / Z class is selected. We can now determine the remaining binary chirality choice. The already identified coefficient has I ( Δ R ) = − X 4 Y 8 I(\Delta_{R}) = -X_{4}Y_{8} I ( Δ R ) = − X 4 Y 8 , whereas the ordered source polynomial is + X 4 Y 8 +X_{4}Y_{8} + X 4 Y 8 [3 equations (74)–(76)] . This nonzero universal polynomial fixes the source coefficient as − Δ R -\Delta_{R} − Δ R in the positive-index convention, yielding
I d i f f ( C ) = χ − Δ R ( C ) = exp ( + π i ξ Δ ( C ) ) = exp ( − 2 π i μ I ( C ) ) . (118) I_{\mathrm{diff}}(C) = \chi_{-\Delta_{R}}(C) = \exp(+\pi i\xi_{\Delta}(C)) = \exp(-2\pi i\mu_{I}(C)). \tag*{(118)} I diff ( C ) = χ − Δ R ( C ) = exp ( + π i ξ Δ ( C )) = exp ( − 2 π i μ I ( C )) . (118)
The polynomial fixes the chirality of this actual system; it does not classify additional flat anomaly sectors. As a separate sign check, the even real APS index in twelve dimensions gives χ R ( ∂ Q ) = exp ( − 2 π i ∫ Q I ( R ) ) \chi_{R}(\partial Q) = \exp(-2\pi i\int_{Q} I(R)) χ R ( ∂ Q ) = exp ( − 2 π i ∫ Q I ( R )) , the inverse bordism phase required by the source’s positively signed consistency law.
For framed trivial gauge connections the reference coefficient is the actual zero virtual real module 256 − 128 − 128 256 - 128 - 128 256 − 128 − 128 , so its difference phase is the unit. On the connected boundary the remaining trivial gauge summands have even real ranks 120, 128 and 256, and hence even chiral kernel multiplicities. These facts are useful parity data for the specified contraction, but they do not prove that all component-ordering or orientation signs in global real gluing vanish.
Finite connection transport of the complete scalar
The closed operator comparison initially uses product-collar Levi–Civita conventions. Write μ L C ( C ) = μ I ( C ; A L C ) \mu_{\mathrm{LC}}(C) = \mu_{I}(C; A_{\mathrm{LC}}) μ LC ( C ) = μ I ( C ; A LC ) for this uncorrected spectral value. The disk calculation uses the stable tangent connection A ∗ A_* A ∗ . Let
α = CS λ ( A L C , A ∗ ) , d α = λ ( A ∗ ) − λ ( A L C ) , H ∗ = H L C + α . (119) \alpha= \operatorname{CS}_{\lambda}(A_{\mathrm{LC}}, A_*), \qquad d\alpha= \lambda(A_*) - \lambda(A_{\mathrm{LC}}), \qquad H_* = H_{\mathrm{LC}} + \alpha. \tag*{(119)} α = CS λ ( A LC , A ∗ ) , d α = λ ( A ∗ ) − λ ( A LC ) , H ∗ = H LC + α . (119)
The last equality transports the existing differential characteristic trivialization along the specified connection homotopy. It retains its integral class, its endpoint isomorphisms and the boundary identification; it is not the selection of an unrelated primitive for the new curvature.
Since Y 8 Y_8 Y 8 depends only on the gauge connections, the tangent transgression of I 12 I_{12} I 12 is α Y 8 \alpha Y_8 α Y 8 . The paths match on the collars, and the reference gauges are trivial, so Y 8 ∣ W r e f = 0 Y_8|_{W_{\mathrm{ref}}}=0 Y 8 ∣ W ref = 0 . The differential-index convention and real APS variation then give
μ ∗ ( C ) − μ L C ( C ) = ∫ C α Y 8 = − ∫ N α Y 8 ( m o d Z ) , (120) \mu_*(C)-\mu_{\mathrm{LC}}(C)=\int_C\alpha Y_8=-\int_N\alpha Y_8\pmod{\mathbb{Z}}, \tag*{(120)} μ ∗ ( C ) − μ LC ( C ) = ∫ C α Y 8 = − ∫ N α Y 8 ( mod Z ) , (120)
a ∗ − a L C = ∫ N α Y 8 , ( a + μ ) ∗ = ( a + μ ) L C ( m o d Z ) . a_*-a_{\mathrm{LC}}=\int_N\alpha Y_8,\qquad(a+\mu)_*=(a+\mu)_{\mathrm{LC}}\pmod{\mathbb{Z}}. a ∗ − a LC = ∫ N α Y 8 , ( a + μ ) ∗ = ( a + μ ) LC ( mod Z ) .
The orientation of N ‾ \overline{N} N fixes the first minus sign. Neither term is separately invariant: their corrections cancel in the complete scalar. On differential String cylinders with trivial gauges, both Y 8 Y_8 Y 8 and Δ R \Delta_{\mathbb{R}} Δ R vanish, so the source law also preserves the scalar with its transported reference.
The transported μ ∗ \mu_* μ ∗ is the differential-index value defined in § 2 , with its finite Chern–Simons correction. It is not being identified with the raw spectrum of a torsionful Dirac operator. This matters even for a topological final answer: using a ∗ a_* a ∗ alongside μ L C \mu_{\mathrm{LC}} μ LC would omit one of the two compensating terms in (120) . Conversely, setting a ∗ = 0 a_*=0 a ∗ = 0 in a chosen geometric representative does not establish that the local term vanishes in every connection convention or on every permitted background.
Substituting (118) into (114) , then applying (120) , gives the identification on the full permitted filling/reference data [7 §9] :
O T Z ( M ) = exp [ − 2 π i ( a ∗ + μ ∗ ( C ) ) ] = O ( M ) , ϕ = − a ∗ − μ ∗ ( m o d Z ) . (121) \mathcal{O}_{\mathrm{TZ}}(M)=\exp\left[-2\pi i(a_*+\mu_*(C))\right]=\mathcal{O}(M),\qquad\phi=-a_*-\mu_*\pmod{\mathbb{Z}}. \tag*{(121)} O TZ ( M ) = exp [ − 2 π i ( a ∗ + μ ∗ ( C )) ] = O ( M ) , ϕ = − a ∗ − μ ∗ ( mod Z ) . (121)
The generally nonzero local integral has not been suppressed. Only after this equality do we use the selected Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) calculation in § 6 , with a ∗ = 0 a_*=0 a ∗ = 0 and μ ∗ = 2 / 3 \mu_*=2/3 μ ∗ = 2/3 and the specified right-frame transport. It gives the known value O T Z = exp ( 2 π i / 3 ) \mathcal{O}_{\mathrm{TZ}}=\exp(2\pi i/3) O TZ = exp ( 2 π i /3 ) [3 §3.2, equation (63)] .
Under the same prescription, Tachikawa–Zhang’s published analysis [3 §4.4] identifies the non-tachyonic fermion spectrum with the first system plus the chirality-flipped second system. It consequently has the same closed eta expression and scalar. This equality concerns the response under the stated spectrum prescription; it does not construct an ultraviolet string theory or identify complete anomaly functors on arbitrary families or nonlifting quotient backgrounds.
The first-root bundle, the effective system and the published slant
The preceding identification uses the full gauge polynomial. It is compatible with distinguishing, rather than interchanging, three specific systems appearing in the published discussion. For the first-root bundle H R ⊕ R 12 \mathbb{H}_{\mathbb{R}} \oplus\mathbb{R}^{12} H R ⊕ R 12 with c = c 2 ( H ) c=c_2(\mathbb{H}) c = c 2 ( H ) , the total Pontryagin class is 1 − 2 c + c 2 1-2c+c^2 1 − 2 c + c 2 . Thus n = − c n=-c n = − c and p 2 = c 2 p_2=c^2 p 2 = c 2 . With the other factor trivial,
Y 8 = n 2 − p 2 6 = c 2 − c 2 6 = 0. (122) Y_8=\frac{n^2-p_2}{6}=\frac{c^2-c^2}{6}=0. \tag*{(122)} Y 8 = 6 n 2 − p 2 = 6 c 2 − c 2 = 0. (122)
This does not give the c 2 / 6 c^2/6 c 2 /6 printed in [3, equation (77)]. The $p_2$ term in their full {xrefext}equation (76)` is indispensable. The two-summand disk bundle used here has E = 2 Q R ⊕ R 8 E=2\mathcal{Q}_{\mathbb{R}}\oplus\mathbb{R}^8 E = 2 Q R ⊕ R 8 , n = − 2 C n=-2C n = − 2 C , p 2 = 6 C 2 p_2=6C^2 p 2 = 6 C 2 and Y 8 = − C 2 / 3 Y_8=-C^2/3 Y 8 = − C 2 /3 , where C = c 2 ( Q , A ) C=c_2(Q,A) C = c 2 ( Q , A ) is the disk curvature form. Its actual connections, boundary frames and integral differential data have been constructed above. The different substitution cannot by itself supply those data.
The seven-dimensional representation printed in [3 equations (57)–(62)] differs from the effective coefficient of the associated-family calculation. Let 4 4 4 and 10 10 10 denote the complex defining and adjoint representations of Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) . For defining roots x , − x , y , − y x,-x,y,-y x , − x , y , − y , write T 4 = 2 ( x 4 + y 4 ) T_4=2(x^4+y^4) T 4 = 2 ( x 4 + y 4 ) . The adjoint roots ± 2 x , ± 2 y , ± x ± y \pm2x,\pm2y,\pm x\pm y ± 2 x , ± 2 y , ± x ± y , together with two zero roots, give
tr 10 F 4 = 12 tr 4 F 4 + 3 ( tr 4 F 2 ) 2 . \operatorname{tr}_{10}F^4=12\operatorname{tr}_4F^4+3\left(\operatorname{tr}_4F^2\right)^2. tr 10 F 4 = 12 tr 4 F 4 + 3 ( tr 4 F 2 ) 2 .
In Chern-character normalization the primitive quartic coefficients are 1 / 24 1/24 1/24 and 12 / 24 12/24 12/24 . The printed system therefore gives
R T Z = − 10 ⋅ 4 + 10 + 30 , − 10 + 12 24 T 4 = 4 48 T 4 . (123) R_{\mathrm{TZ}}=-10\cdot4+10+30,\qquad\frac{-10+12}{24}T_4=\frac{4}{48}T_4. \tag*{(123)} R TZ = − 10 ⋅ 4 + 10 + 30 , 24 − 10 + 12 T 4 = 48 4 T 4 . (123)
The representation-valued fibre index of the present actual-bundle construction gives R e f f = − 4 ⋅ 4 + 16 R_{\mathrm{eff}}=-4\cdot4+16 R eff = − 4 ⋅ 4 + 16 , whose primitive coefficient is − 8 T 4 / 48 -8T_4/48 − 8 T 4 /48 [7 Appendix B] . The qualifier “printed” is important here. Wada–Yonekura instead use R ~ a u x = 10 ⋅ 4 − 10 − 30 ⋅ 1 \widetilde{R}_{\mathrm{aux}}=10\cdot4-10-30\cdot1 R aux = 10 ⋅ 4 − 10 − 30 ⋅ 1 [6 Section 9] . With the trace normalization above, its primitive quartic coefficient is
10 − 12 24 T 4 = − 4 48 T 4 . \frac{10-12}{24}T_4=-\frac{4}{48}T_4. 24 10 − 12 T 4 = − 48 4 T 4 .
which is half the primitive coefficient of R e f f R_{\mathrm{eff}} R eff . This comparison of coefficients is not an identity of virtual representations. More precisely, in their specified S 3 S^3 S 3 construction Wada–Yonekura establish an equality of index densities and hence the index relation [6 equations (9.13)–(9.14)]
Ind C ( D S 8 + ⊗ R ~ a u x ) = 1 2 Ind C ( D S 8 + ⊗ V g ′ ) . \operatorname{Ind}_{\mathbb{C}}\left(D^+_{S^8}\otimes\widetilde{R}_{\mathrm{aux}}\right)=\frac{1}{2}\operatorname{Ind}_{\mathbb{C}}\left(D^+_{S^8}\otimes V_{g'}\right). Ind C ( D S 8 + ⊗ R aux ) = 2 1 Ind C ( D S 8 + ⊗ V g ′ ) .
where the bundles and the notation V g ′ V_{g'} V g ′ are those of [6 ] . We do not identify V g ′ V_{g'} V g ′ with R e f f R_{\mathrm{eff}} R eff as a virtual bundle or assert an exact dictionary between the intermediate geometric and spectral data of the two approaches. In particular, the index relation does not identify their eta lifts, connections, framings or differential fillings. Comparisons of signed scalar values also require matching orientation and phase conventions.
The full effective polynomial also keeps more information than that primitive coefficient. In the source normalization it is
I ~ 8 , e f f = − 1 6 T 4 − 1 6 p 1 c 2 . \widetilde{I}_{8,\mathrm{eff}}=-\frac{1}{6}T_4-\frac{1}{6}p_1c_2. I 8 , eff = − 6 1 T 4 − 6 1 p 1 c 2 .
The representation has rank zero, so the pure gravitational term cancels. On the stabilizer K = Sp ( 1 ) K = \operatorname{Sp}(1) K = Sp ( 1 ) , the restriction 4 ∣ K = U + 2 4|_{K} = U + 2 4 ∣ K = U + 2 gives R e f f ∣ K = − 4 U + 8 R_{\mathrm{eff}}|_{K} = -4U + 8 R eff ∣ K = − 4 U + 8 and T 4 = 2 c 2 T_{4} = 2c^{2} T 4 = 2 c 2 , where U U U is the defining complex representation of K K K and c = c 2 ( U ) c = c_{2}(U) c = c 2 ( U ) . With p 1 = 2 λ p_{1} = 2\lambda p 1 = 2 λ , this yields I 8 , e f f = − ( λ + c ) c / 3 I_{8,\mathrm{eff}} = -(\lambda+ c)c/3 I 8 , eff = − ( λ + c ) c /3 . Thus the auxiliary factor used below comes from a specified representation and its associated connection, not from selecting a factorization that resembles the published brane polynomial.
The published slant construction provides another way to relate a seven-dimensional base to an eleven-dimensional relative anomaly evaluation [3 equations (A.16)–(A.23)] . Comparing its domain with the present disk construction is necessary before using such an evaluation for the same response: a shared dimension or third-root value does not identify the gauge twist or the differential filling.
Here M S p i n M\mathrm{Spin} M Spin and M S t r i n g M\mathrm{String} M String are the spectra representing spin and String bordism: stable objects packaging bordism in all dimensions. Their homotopy groups record manifolds with the corresponding structures up to bordism. The map M S t r i n g → M S p i n M\mathrm{String} \to M\mathrm{Spin} M String → M Spin forgets the String trivialization. Its cofiber C 0 = cofib ( M S t r i n g → M S p i n ) C_{0} = \operatorname{cofib}(M\mathrm{String} \to M\mathrm{Spin}) C 0 = cofib ( M String → M Spin ) represents relative bordism: a spin filling with a specified String structure on its boundary defines a relative class. In particular, π 11 C 0 \pi_{11}C_{0} π 11 C 0 contains classes of eleven-dimensional spin fillings of ten-dimensional String manifolds. The space B S U ( 2 ) B\mathrm{SU}(2) B SU ( 2 ) classifies the principal S U ( 2 ) \mathrm{SU}(2) SU ( 2 ) gauge bundle; the added twist + 2 c +2c + 2 c specifies a trivialization of λ − 2 c \lambda- 2c λ − 2 c .
The notation I Z I_{\mathbb{Z}} I Z denotes integral Anderson duality. For these bordism spectra it supplies anomaly classes with an evaluation on bordism cycles, including the torsion pairing valued in Q / Z \mathbb{Q}/\mathbb{Z} Q / Z that curvature forms alone cannot detect. A torsion class ζ \zeta ζ in degree twelve of I Z C 0 I_{\mathbb{Z}}C_{0} I Z C 0 pairs with a torsion relative cycle in degree eleven, as in [3 equations (A.16)–(A.17)] . The published slant class extends the four-dimensional relative disk class [ D 4 , S 3 ] [D^{4}, S^{3}] [ D 4 , S 3 ] equivariantly, with the Lie-group-framed String structure on its boundary. Its cohomological degree − 4 -4 − 4 records this four-dimensional shift: acting on a seven-dimensional twisted String cycle b b b produces an eleven-dimensional relative cycle. The induced dual map I Z a s l a n t I_{\mathbb{Z}}a_{\mathrm{slant}} I Z a slant precomposes an anomaly evaluation with that slant map. In these conventions,
a s l a n t ∈ C 0 − 4 + 2 c ( B S U ( 2 ) ) , ( a s l a n t ) ∗ : M S t r i n g 7 + 2 c ( B S U ( 2 ) ) ⟶ π 11 C 0 , a_{\mathrm{slant}} \in C_{0}^{-4+2c}(B\mathrm{SU}(2)), \qquad(a_{\mathrm{slant}})_{*}: M\mathrm{String}_{7+2c}(B\mathrm{SU}(2)) \longrightarrow\pi_{11}C_{0}, a slant ∈ C 0 − 4 + 2 c ( B SU ( 2 )) , ( a slant ) ∗ : M String 7 + 2 c ( B SU ( 2 )) ⟶ π 11 C 0 ,
⟨ ζ , ( a s l a n t ) ∗ b ⟩ = ⟨ ( I Z a s l a n t ) ( ζ ) , b ⟩ . (124) \langle\zeta, (a_{\mathrm{slant}})_{*}b\rangle= \langle(I_{\mathbb{Z}}a_{\mathrm{slant})(\zeta), b\rangle.} \tag*{(124)} ⟨ ζ , ( a slant ) ∗ b ⟩ = ⟨( I Z a slant ) ( ζ ) , b ⟩ . (124)
The last equality is Anderson evaluation naturality for compatible classes and cycles, with the torsion hypotheses when its values lie in Q / Z \mathbb{Q}/\mathbb{Z} Q / Z . It is a formal adjunction, rather than an identification of the actual-bundle gauge and differential data with the published instanton datum. The published + 2 c +2c + 2 c domain imposes λ = 2 c \lambda= 2c λ = 2 c . Substitution into the auxiliary seven-dimensional polynomial would instead give
− 1 3 ( λ + c ) c ∣ λ = 2 c = − c 2 , -\frac{1}{3}(\lambda+ c)c\big|_{\lambda=2c} = -c^{2}, − 3 1 ( λ + c ) c λ = 2 c = − c 2 ,
which does not vanish. The auxiliary factor λ + c \lambda+ c λ + c , the actual eleven-dimensional twist t t w = − 2 c t_{\mathrm{tw}} = -2c t tw = − 2 c , and the published + 2 c +2c + 2 c slant thus have different roles. The integral TMF and vertex operator superalgebra (VOSA) discussion that follows requires its own compatible lifts and selected differential filling; neither the ordered scalar identification nor a matching third-root number constructs those additional inputs.
Conditional integral TMF and VOSA consequences
The calibrated response on Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) can test a supplied integral class in topological modular forms (TMF). Here TMF is a generalized cohomology theory whose integral groups retain torsion information beyond the modular forms seen rationally [21 §1.3] . This is a different use of the anomaly calculation from the published fermion comparison in § 9 . The normalized scalar is defined without a vertex operator superalgebra (VOSA) assignment. An assignment to TMF would instead let us ask which torsion class is detected by that scalar. The answer below is conditional on compatible integral lifts with their real equivariant characters; the character calculation does not construct the lifts. We explain the selected comparison of Li–Sun [7 §10] , using the relative Anderson construction of Tachikawa–Yamashita in its fixed version 3 [21 ] .
The broader motivation is the Stolz–Teichner proposal that concordance classes of local 2 ∣ 1 2|1 2∣1 -dimensional Euclidean field theories realize TMF [22 Conjecture 1.17] . These are two-dimensional supersymmetric theories in their geometric formulation. A modular-form-valued elliptic index does not retain all of the proposed integral information. Gaiotto–Johnson-Freyd’s secondary elliptic-genus invariant detects some torsion, but vanishes for the proposed supersymmetric quantum field theory associated with the Lie-framed Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) class [23 §1, §2.5] . The relative Anderson pairing below tests a supplied integral class against this degree-ten background. It supplies a torsion detector under the stated lift hypotheses, without establishing the proposed field-theory description of TMF.
The relevant test pairs a degree-− 32 -32 − 32 class with the String-oriented image of the ten-dimensional, left Lie-framed Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) . Two ingredients make this test computable. A constant-coefficient functional selects exactly the virtual real fermion coefficient already used to define the response. The selected differential disk filling of § 6 has zero local term, so its relative Anderson evaluation reduces to a closed spin/gauge evaluation. Both ingredients are needed: the representation identity determines the closed coefficient, while the differential filling determines whether an additional relative contribution remains.
Integral lifts and the eleven-dimensional twist
The bordism notation records which geometric data may extend across a filling. The spectra MSpin and MString represent spin and String bordism: their groups in dimension n n n consist of closed n n n -manifolds with the indicated structure, identified when joined by a bordism preserving that structure. A String structure includes an integral trivialization of λ \lambda λ . A spectrum is the stable object that collects these groups and their operations across dimensions; π n \pi_n π n denotes its degree-n n n group. The map MString → \to → MSpin forgets the integral String trivialization. Its cofiber C 0 C_0 C 0 represents relative spin/String bordism, in which a spin filling carries a specified String structure on its boundary [21 §2.3, Appendix E] . This relative object retains the boundary datum even when the boundary already bounds as a spin manifold. It is therefore suited to evaluating the globally normalized response rather than only its local polynomial.
The other coefficient theory used below is real K K K -theory, KO, whose degree-zero classes are virtual real vector bundles. K O ( ( q ) ) \mathrm{KO}((q)) KO (( q )) allows formal Laurent series in q q q with those KO coefficients. The Tate-character map from TMF to K O ( ( q ) ) \mathrm{KO}((q)) KO (( q )) extracts the real bundle-valued character series, retaining a gauge action when working equivariantly [21 §1.3, §4.5] . Forgetting that action gives a numerical character with less information. We use these theories only through their stated character map, orientations and relative pairings.
Keep K = S p ( 1 ) K = \mathrm{Sp}(1) K = Sp ( 1 ) , with B K BK B K the space classifying principal K K K -bundles, and let c = c 2 ∈ H 4 ( B K ; Z ) c = c_2 \in H^4(BK;\mathbb{Z}) c = c 2 ∈ H 4 ( B K ; Z ) denote its universal class. The notation B K + BK_+ B K + adds a disjoint basepoint. Smashing a bordism spectrum with it adjoins a gauge background, while the subscript t t w \mathrm{ttw} ttw imposes the twisted String trivialization. We denote spin/gauge bordism by B t t w B_{\mathrm{ttw}} B ttw , its twisted String refinement by S t t w S_{\mathrm{ttw}} S ttw , and their relative bordism theory by C t t w C_{\mathrm{ttw}} C ttw [21 Appendix E] . In this section the twist is
t t w = − 2 c , S t t w = M S t r i n g ∧ t t w B K + , t_{\mathrm{tw}} = -2c,\qquad S_{\mathrm{ttw}} = \mathrm{MString} \mathbin{\wedge}_{\mathrm{ttw}} BK_+, t tw = − 2 c , S ttw = MString ∧ ttw B K + ,
B t t w = M S p i n ∧ B K + , C t t w = cofib ( S t t w ⟶ B t t w ) . (125) B_{\mathrm{ttw}} = \mathrm{MSpin} \wedge BK_+,\qquad C_{\mathrm{ttw}} = \operatorname{cofib}(S_{\mathrm{ttw}} \longrightarrow B_{\mathrm{ttw}}). \tag*{(125)} B ttw = MSpin ∧ B K + , C ttw = cofib ( S ttw ⟶ B ttw ) . (125)
A t t w t_{\mathrm{tw}} t tw -twisted String structure trivializes λ − t t w = λ + 2 c \lambda- t_{\mathrm{tw}} = \lambda+ 2c λ − t tw = λ + 2 c . This is the eleven-dimensional gauge twist supplied by λ ( E L ) + λ ( E L ′ ) = − 2 c \lambda(E_L) + \lambda(E'_L) = -2c λ ( E L ) + λ ( E L ′ ) = − 2 c . It differs from the seven-dimensional base trivialization of λ + c \lambda+ c λ + c used to construct the disk datum. It also differs from the Crowley–Goette invariant t B ( V ) t_B(V) t B ( V ) : t t w t_{\mathrm{tw}} t tw names a degree-four integral twist, whereas t B ( V ) t_B(V) t B ( V ) is a secondary invariant with its separate domain.
Write C 0 = cofib ( M S t r i n g → M S p i n ) C_0 = \operatorname{cofib}(\mathrm{MString} \to\mathrm{MSpin}) C 0 = cofib ( MString → MSpin ) . The basepoint of B K BK B K induces i : C 0 → C t t w i : C_0 \to C_{\mathrm{ttw}} i : C 0 → C ttw , and the cofiber map is j : B t t w → C t t w j : B_{\mathrm{ttw}} \to C_{\mathrm{ttw}} j : B ttw → C ttw . Geometrically, i i i equips a relative cycle with the trivial gauge background, while j j j regards a closed spin/gauge cycle as a relative cycle with empty boundary. With M = S p ( 2 ) M = \mathrm{Sp}(2) M = Sp ( 2 ) carrying s L s_L s L and the same trivial-gauge spin reference W r e f W_{\mathrm{ref}} W ref , put
z = [ W r e f , M ] ∈ π 11 C 0 , Z = N ‾ ∪ M W r e f . (126) z = [W_{\mathrm{ref}}, M] \in\pi_{11}C_0,\qquad Z = \overline{N} \cup_M W_{\mathrm{ref}}. \tag*{(126)} z = [ W ref , M ] ∈ π 11 C 0 , Z = N ∪ M W ref . (126)
Here N N N and W r e f W_{\mathrm{ref}} W ref are eleven-dimensional fillings. The base filling in the relative characteristic calculation is eight-dimensional, and the spin/gauge filling in the APS calculation is twelve-dimensional. Their dimensions and roles differ from those of the relative cycle above. The orientation of the closed double reverses the disk side, as in the earlier evaluation of μ \mu μ .
Let Φ \Phi Φ denote the Tate-character map. For any degree n n n , the group
A n = ker ( Φ : π n T M F ⟶ π n K O ( ( q ) ) ) A_n = \ker(\Phi: \pi_n\mathrm{TMF} \longrightarrow\pi_n\mathrm{KO}((q))) A n = ker ( Φ : π n TMF ⟶ π n KO (( q )))
consists of classes invisible to the character. In particular A 10 A_{10} A 10 , used below, is the degree-ten group containing the String-oriented test manifold’s image. A superscript 32 + t t w 32 + t_{\mathrm{tw}} 32 + t tw denotes cohomological degree 32 with the integral twist, and restriction to a point has homotopy degree − 32 -32 − 32 . The proposed assignments must carry the gauge-equivariant data needed to relate these two degrees through the relative evaluation.
Hypothesis 10.1 (Compatible integral lifts). There are classes u D , u E ∈ T M F 32 + t t w ( B K ) u_D,u_E \in\mathrm{TMF}^{32+t_{\mathrm{tw}}}(BK) u D , u E ∈ TMF 32 + t tw ( B K ) with compatible real equivariant Tate characters. Their restrictions along p t → B K \mathrm{pt} \to BK pt → B K are x D , x E x_D,x_E x D , x E , and x = x D − x E x = x_D - x_E x = x D − x E belongs to A − 32 A_{-32} A − 32 . The subscripts denote the rank-sixteen lattice inputs associated to Spin ( 32 ) / Z 2 \operatorname{Spin}(32)/\mathbb{Z}_2 Spin ( 32 ) / Z 2 and E 8 × E 8 E_8 \times E_8 E 8 × E 8 , in that order. Their characters have the real parity normalization of published Tachikawa–Zhang (A.13) [3 (A.13)] . If a further
VOSA is compared, its assignment supplies an actual u V ∈ T M F 32 + t w ( B K ) u_V \in\mathrm{TMF}^{32+tw}(BK) u V ∈ TMF 32 + tw ( B K ) with the same equivariant K O ( ( q ) ) \mathrm{KO}((q)) KO (( q )) image as u D − u E u_D-u_E u D − u E .
These are hypotheses about integral classes, restriction maps and real representation-valued characters. Equality of numerical q q q -series after forgetting the gauge action would not provide them. In the published comparison, the VOSA assignments are formulated as conjectures [3 Conjectures A.7 and A.12] . We use an assignment only when it supplies the stated class, and do not infer it from the existence of a lattice algebra or from agreement of local anomaly polynomials.
The real coefficient selected by the character series
Under Hypothesis 10.1, the common leading q − 2 q^{-2} q − 2 terms of the two lattice inputs cancel. The next term is the difference of their gauge adjoints. On the common Spin ( 16 ) × Spin ( 16 ) \operatorname{Spin}(16) \times\operatorname{Spin}(16) Spin ( 16 ) × Spin ( 16 ) cover, the actual representation decompositions are
Λ 2 ( E ⊕ E ′ ) = Λ 2 E ⊕ ( E ⊗ E ′ ) ⊕ Λ 2 E ′ , e 8 ∣ Spin ( 16 ) = Λ 2 E ⊕ S R + ( E ) . (127) \begin{aligned}
\Lambda^2(E \oplus E') &= \Lambda^2 E \oplus(E \otimes E') \oplus\Lambda^2 E', \\
\mathfrak{e}_8|_{\operatorname{Spin}(16)} &= \Lambda^2 E \oplus S_{\mathrm{R}}^+(E). \tag*{(127)}
\end{aligned} Λ 2 ( E ⊕ E ′ ) e 8 ∣ Spin ( 16 ) = Λ 2 E ⊕ ( E ⊗ E ′ ) ⊕ Λ 2 E ′ , = Λ 2 E ⊕ S R + ( E ) . (127)
Subtracting the two E 8 E_8 E 8 adjoints cancels both exterior squares. The remaining virtual real coefficient is therefore
Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , Φ ( u D − u E ) = q − 1 Δ R + O ( q 0 ) . (128) \Delta_{\mathrm{R}} = E \otimes_{\mathbb{R}} E' - S_{\mathrm{R}}^+(E) - S_{\mathrm{R}}^+(E'), \qquad\Phi(u_D-u_E) = q^{-1}\Delta_{\mathrm{R}} + O(q^0). \tag*{(128)} Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) , Φ ( u D − u E ) = q − 1 Δ R + O ( q 0 ) . (128)
The real structures and Ramond parity in the hypothesis fix this sign and normalization. This identity retains the representations, rather than replacing them by their ranks. It is the coefficient identity used by Li–Sun [7 Lemma 10.2] , and in the published character comparison [3 (A.13)] .
Distinguish the modular discriminant Δ m o d ( q ) = q + O ( q 2 ) \Delta_{\mathrm{mod}}(q) = q + O(q^2) Δ mod ( q ) = q + O ( q 2 ) from the virtual coefficient Δ R \Delta_{\mathrm{R}} Δ R . The Witten bundle series satisfies Wit ( q ) = 1 + O ( q ) \operatorname{Wit}(q) = 1 + O(q) Wit ( q ) = 1 + O ( q ) . Thus
[ Δ m o d ( q ) Wit ( q ) Φ ( u D − u E ) ] q 0 = Δ R . (129) [\Delta_{\mathrm{mod}}(q)\operatorname{Wit}(q)\Phi(u_D-u_E)]_{q^0} = \Delta_{\mathrm{R}}. \tag*{(129)} [ Δ mod ( q ) Wit ( q ) Φ ( u D − u E ) ] q 0 = Δ R . (129)
This extraction is exact. After the q − 2 q^{-2} q − 2 cancellation, the character series has lowest power q − 1 q^{-1} q − 1 , while the product of the discriminant and Witten series has lowest power q q q . Their leading terms are the only terms that can produce q 0 q^0 q 0 . Higher q q q -coefficients consequently do not enter this particular closed evaluation; this says nothing about equality of the full theories or about other character-dependent observables.
The selected differential term and the positive pairing
Anderson duality supplies the evaluation that detects integral torsion. For a spectrum F F F , its Anderson dual I Z F I_{\mathbb{Z}}F I Z F records integer-valued homomorphisms on its coefficient groups together with extension data from torsion. In degree r r r , that torsion part can evaluate a torsion class of degree r − 1 r-1 r − 1 in Q / Z \mathbb{Q}/\mathbb{Z} Q / Z ; exponentiation turns the result into a finite response phase. This is why the relative evaluation can detect a class that has zero Tate character, rather than merely reproduce a curvature integral [21 §1.7, Definition 3.48] .
The untwisted relative Anderson map of Tachikawa–Yamashita is α s p i n / s t r i n g : T M F → Σ − 20 I Z C 0 \alpha_{\mathrm{spin/string}}:\mathrm{TMF}\to\Sigma^{-20}I_{\mathbb{Z}}C_0 α spin/string : TMF → Σ − 20 I Z C 0 . Its degree-− 32 -32 − 32 input determines an Anderson class in degree twelve, which pairs with the torsion relative cycle z z z in degree eleven. The suspension Σ − 20 \Sigma^{-20} Σ − 20 shifts the degree of the dual by twenty; cohomological degree twelve is consequently the degree that evaluates this eleven-dimensional relative cycle. The String orientation σ : M S t r i n g → T M F \sigma:\mathrm{MString}\to\mathrm{TMF} σ : MString → TMF sends a String bordism class to its associated TMF class. Applied to [ M , s L ] [M,s_L] [ M , s L ] , it gives the ten-dimensional class against which x x x is tested. The relative torsion evaluation is precisely ( ( x , σ [ M , s L ] ) ) ((x,\sigma[M,s_L])) (( x , σ [ M , s L ])) by Tachikawa–Yamashita [21 Definition 3.48 and Proposition 3.55] . The general differential construction can use the underlying integral class when H 11 ( C t t w ; R ) = 0 H^{11}(C_{t_{\mathrm{tw}}};\mathbb{R})=0 H 11 ( C t tw ; R ) = 0 [21 §3.3.1] . The rational calculation below establishes this vanishing, so no additional choice of differential TMF lift is needed here.
Naturality of the twisted construction separates the evaluation into the closed double and the extending differential filling:
i ∗ z = j ∗ [ Z ] + [ N , M ] d i f f . (130) i_*z=j_*[Z]+[N,M]_{\mathrm{diff}}. \tag*{(130)} i ∗ z = j ∗ [ Z ] + [ N , M ] diff . (130)
The last term retains the extending t t w t_{\mathrm{tw}} t tw -twisted differential String structure, gauge connection, product collars and boundary identifications. Its underlying relative class is nullbordant, but this alone would not set its differential evaluation to zero. Differential information is precisely what can distinguish refinements of a topologically null class.
For the selected filling, that evaluation can be computed using the cylinder N × [ 0 , 1 ] N\times[0,1] N × [ 0 , 1 ] with the intervals over M M M collapsed to the relative corner and smoothly rounded. One face carries the spin data and the other its extending twisted String structure; the fixed collars identify their common boundary. Gauge data are pulled back and tangent data are stably product, including the connection transgressions at the corner. In the mapping-cone characteristic formula, the degree-twelve bulk form is pulled back from the eleven-dimensional N N N and vanishes by dimension. Write Y r e l Y_{\mathrm{rel}} Y rel for the factor in the relative characteristic form of Tachikawa–Yamashita. The remaining String-face integral uses
Y r e l = − Y 8 = 1 3 C 2 , ∫ N H L Y r e l = − ∫ N H L Y 8 = − a N , ∗ = 0. (131) Y_{\mathrm{rel}}=-Y_8=\frac{1}{3}C^2,\qquad\int_N H_LY_{\mathrm{rel}}=-\int_N H_LY_8=-a_{N,*}=0. \tag*{(131)} Y rel = − Y 8 = 3 1 C 2 , ∫ N H L Y rel = − ∫ N H L Y 8 = − a N , ∗ = 0. (131)
The last equality is the exact local calculation at the connection of § 6 . There the three-form is H L = 2 T q + K 3 H_L=2T_q+K_3 H L = 2 T q + K 3 because the base curving is h = 0 h=0 h = 0 , while its integral refinement is retained. Both fibre integrals π ∗ ( T q C 2 ) \pi_*(T_qC^2) π ∗ ( T q C 2 ) and π ∗ ( K 3 C 2 ) \pi_*(K_3C^2) π ∗ ( K 3 C 2 ) vanish as forms, so the String-face integral is zero at these actual connections. The mapping-cone construction of Tachikawa–Yamashita is described in [21 Appendix E, (E.12)–(E.14) and (E.18)] ; Li–Sun [7 Lemma 10.3] supply this selected application. It proves zero evaluation of this term, without asserting that every differential refinement of the underlying null class is zero. Additivity in (130) now reduces the pairing to the closed evaluation on Z Z Z .
For degree 32 32 32 and a closed eleven-dimensional spin manifold, the real differential KO pushforward has period κ 12 Z ( ( q ) ) = 2 Z ( ( q ) ) \kappa_{12}\mathbb{Z}((q))=2\mathbb{Z}((q)) κ 12 Z (( q )) = 2 Z (( q )) . The positive functional of Tachikawa–Yamashita is
f ( q ) ⟼ 1 2 [ Δ m o d ( q ) f ( q ) ] q 0 ( m o d Z ) , (132) f(q)\longmapsto\frac{1}{2}\left[\Delta_{\mathrm{mod}}(q)f(q)\right]_{q^0}\pmod{\mathbb{Z}}, \tag*{(132)} f ( q ) ⟼ 2 1 [ Δ mod ( q ) f ( q ) ] q 0 ( mod Z ) , (132)
with the real-period interpretation of their equation (3.95), and the functional in (3.99) and Proposition 3.100 [21 ] . Its differential relative target is the quotient
R ( ( q ) ) / ( ( M F − 10 ) R + 2 Z ( ( q ) ) ) . \mathbb{R}((q))/((\mathrm{MF}_{-10})_{\mathbb{R}}+2\mathbb{Z}((q))). R (( q )) / (( MF − 10 ) R + 2 Z (( q ))) .
where ( M F − 10 ) R (\mathrm{MF}_{-10})_{\mathbb{R}} ( MF − 10 ) R denotes the real modular-form ambiguity. Multiplication by the weight-twelve discriminant sends that ambiguity to weight two, whose constant coefficient vanishes [21 Fact 2.8] ; the functional therefore descends to the quotient. The factor 1 / 2 1/2 1/2 acts on this real coefficient quotient: a change by an even APS period changes the output by an integer. It is not division of an arbitrary R / Z \mathbb{R}/\mathbb{Z} R / Z eta class by two. In particular, the reduced eta invariant retains its kernel contribution, ξ ( D ) = ( η A P S ( D ) + dim C ker D ) / 2 \xi(D)=(\eta_{\mathrm{APS}}(D)+\dim_{\mathbb{C}}\ker D)/2 ξ ( D ) = ( η APS ( D ) + dim C ker D ) /2 .
Applying (129) to this functional at the Levi–Civita tangent connection A L C A_{\mathrm{LC}} A LC yields the raw spectral value + 1 2 ξ ( D Z , Δ C L C ) = − μ I ( Z ; A L C ) +\frac{1}{2}\xi(D^{\mathrm{LC}}_{Z,\Delta_{\mathbb{C}}})=-\mu_I(Z;A_{\mathrm{LC}}) + 2 1 ξ ( D Z , Δ C LC ) = − μ I ( Z ; A LC ) , where Δ C = Δ R ⊗ R C \Delta_{\mathbb{C}}=\Delta_{\mathbb{R}}\otimes_{\mathbb{R}}\mathbb{C} Δ C = Δ R ⊗ R C . To evaluate the selected stable tangent connection A T A_T A T , we then apply the finite transgression. The polynomial of the positive term is − I 12 -I_{12} − I 12 , so its Chern–Simons correction is the negative of the correction to μ I \mu_I μ I . With the convention of (13) , the corrected value is
+ 1 2 ξ ( D Z , Δ C L C ) − ∫ Z T I 12 ( A L C , A T ) = − μ ∗ ( Z ; A T ) ( m o d Z ) . (133) +\frac{1}{2}\xi(D^{\mathrm{LC}}_{Z,\Delta_{\mathbb{C}}})-\int_Z T_{I_{12}}(A_{\mathrm{LC}},A_T)=-\mu_*(Z;A_T)\pmod{\mathbb{Z}}. \tag*{(133)} + 2 1 ξ ( D Z , Δ C LC ) − ∫ Z T I 12 ( A LC , A T ) = − μ ∗ ( Z ; A T ) ( mod Z ) . (133)
The raw eta value and this single finite correction are separate steps; the stable connection is represented by μ ∗ \mu_* μ ∗ . The selected calibration then gives
( ( x , σ [ M , s L ] ) ) = − μ ∗ ( Z ) = − 2 3 = 1 3 in R / Z , e 2 π i ( ( x , σ [ M , s L ] ) ) = e 2 π i / 3 . (134) ((x,\sigma[M,s_L]))=-\mu_*(Z)=-\frac{2}{3}=\frac{1}{3}\quad\text{in }\mathbb{R}/\mathbb{Z},\qquad e^{2\pi i((x,\sigma[M,s_L]))}=e^{2\pi i/3}. \tag*{(134)} (( x , σ [ M , s L ])) = − μ ∗ ( Z ) = − 3 2 = 3 1 in R / Z , e 2 π i (( x , σ [ M , s L ])) = e 2 π i /3 . (134)
For these data it agrees with the normalized scalar ϕ = − a N , ∗ − μ ∗ \phi=-a_{N,*}-\mu_* ϕ = − a N , ∗ − μ ∗ because a N , ∗ = 0 a_{N,*}=0 a N , ∗ = 0 . This equality uses the selected filling and its finite corrections, as proved by Li–Sun [7 Proposition 10.4] . The generally nonzero local term of a different differential filling has not been discarded, and this calculation does not supply a general comparison formula with that term present.
Two source qualifications matter here. Tachikawa–Yamashita version 3, equation (3.95), prints reduction modulo Z ( ( q ) ) \mathbb{Z}((q)) Z (( q )) while displaying a target with period κ n + 1 Z ( ( q ) ) \kappa_{n+1}\mathbb{Z}((q)) κ n + 1 Z (( q )) . In the present real degree, κ 12 = 2 \kappa_{12}=2 κ 12 = 2 by their equation (3.75); the interpretation above retains that real period and the corresponding APS parity [21 (3.75), (3.95)] . Also, their Proposition 4.59 assumes cohomological degree d ≡ 4 ( m o d 8 ) d\equiv4\pmod{8} d ≡ 4 ( mod 8 ) , which excludes degree 32. The present comparison uses the underlying differential relative construction, the applicable closed formula and the explicit zero-term calculation, rather than extending that proposition’s range. The suspension − 20 -20 − 20 fixes the relative-cycle dimension as 32 − 20 − 1 = 11 32-20-1=11 32 − 20 − 1 = 11 and the curvature and bordism degree as twelve. The general boundary Witten-eta assertion, Conjecture 3.108 in that version, is unused [21 Proposition 4.59 and Conjecture 3.108] . Li–Sun make these delimitations explicit [7 Remark 10.5] .
Restricted Tate detection and the order-three class
The result can identify a further assigned VOSA class because equality of equivariant Tate images determines the evaluation on this particular relative cycle. The selected twisted String nullbordism is essential. It permits the underlying i ∗ z i_*z i ∗ z to come from a closed spin/gauge class through j j j ; an arbitrary relative cycle need not have this property.
Let α t w \alpha_{\mathrm{tw}} α tw denote the twisted counterpart of the relative Anderson map defined above. For two integral lifts with the same equivariant Tate image, let v v v be their difference and γ = α t w ( v ) ∈ I Z 12 ( C t w ) \gamma= \alpha_{\mathrm{tw}}(v) \in I_{\mathbb{Z}}^{12}(C_{\mathrm{tw}}) γ = α tw ( v ) ∈ I Z 12 ( C tw ) . Compatibility with the Tate map gives j ∗ γ = 0 j^*\gamma= 0 j ∗ γ = 0 . To see why γ \gamma γ is torsion, first consider the rational characteristic ring Q [ p 1 , p 2 , … , c ] \mathbb{Q}[p_1,p_2,\ldots,c] Q [ p 1 , p 2 , … , c ] of B t w B_{\mathrm{tw}} B tw . The twisted String condition gives its quotient by p 1 / 2 + 2 c p_1/2 + 2c p 1 /2 + 2 c for S t w S_{\mathrm{tw}} S tw . In the rational model of the homotopy fibre, the added degree-three generator has differential p 1 / 2 + 2 c p_1/2 + 2c p 1 /2 + 2 c . Since this polynomial is a non-zero-divisor, the quotient has only even-degree cohomology, and the map from the spin/gauge ring onto it is surjective. Finite-type rational duality makes the map on even homology injective. The cofiber sequence then implies that π 11 B t w \pi_{11}B_{\mathrm{tw}} π 11 B tw and π 11 C t w \pi_{11}C_{\mathrm{tw}} π 11 C tw are finite torsion groups. It also gives an injection
H 12 ( C t w ; Q ) ⟶ H 12 ( B t w ; Q ) , H^{12}(C_{\mathrm{tw}};\mathbb{Q}) \longrightarrow H^{12}(B_{\mathrm{tw}};\mathbb{Q}), H 12 ( C tw ; Q ) ⟶ H 12 ( B tw ; Q ) ,
because H 11 ( S t w ; Q ) = 0 H^{11}(S_{\mathrm{tw}};\mathbb{Q}) = 0 H 11 ( S tw ; Q ) = 0 . Thus j ∗ γ = 0 j^*\gamma= 0 j ∗ γ = 0 forces the rational image of γ \gamma γ to vanish. The Anderson universal coefficient sequence
0 ⟶ Ext 1 ( π 11 C t w , Z ) ⟶ I Z 12 ( C t w ) ⟶ Hom ( π 12 C t w , Z ) ⟶ 0 (135) 0 \longrightarrow\operatorname{Ext}^{1}(\pi_{11}C_{\mathrm{tw}},\mathbb{Z}) \longrightarrow I_{\mathbb{Z}}^{12}(C_{\mathrm{tw}}) \longrightarrow\operatorname{Hom}(\pi_{12}C_{\mathrm{tw}},\mathbb{Z}) \longrightarrow0 \tag*{(135)} 0 ⟶ Ext 1 ( π 11 C tw , Z ) ⟶ I Z 12 ( C tw ) ⟶ Hom ( π 12 C tw , Z ) ⟶ 0 (135)
places γ \gamma γ in the finite torsion term on the left, while retaining the integer-valued homomorphisms on the right. This identifies the part of Anderson duality that can contribute to the next evaluation. The full rational and universal coefficient proof is given by Li–Sun [7 Lemma 10.6] .
Since the selected nullbordism gives i ∗ z = j ∗ w i_*z = j_*w i ∗ z = j ∗ w for w ∈ π 11 B t w w \in\pi_{11}B_{\mathrm{tw}} w ∈ π 11 B tw , naturality of the torsion pairing gives
⟨ γ , i ∗ z ⟩ = ⟨ γ , j ∗ w ⟩ = ⟨ j ∗ γ , w ⟩ = 0. (136) \langle\gamma,i_*z\rangle= \langle\gamma,j_*w\rangle= \langle j^*\gamma,w\rangle= 0. \tag*{(136)} ⟨ γ , i ∗ z ⟩ = ⟨ γ , j ∗ w ⟩ = ⟨ j ∗ γ , w ⟩ = 0. (136)
Thus the two lifts have equal relative evaluations on the selected class. This is a statement about one pairing with its nullbordism hypothesis, not a claim that the Tate character is injective or determines every relative anomaly evaluation.
Tachikawa–Zhang [3 (A.3)–(A.5)] , and Tachikawa–Yamashita [21 Proposition 3.49 and Remark 3.56] , identify
A 10 = A − 32 = Z / 3 , ( ( − , − ) ) : A − 32 × A 10 ⟶ Q / Z (137) A_{10} = A_{-32} = \mathbb{Z}/3,\qquad((-, -)): A_{-32} \times A_{10} \longrightarrow\mathbb{Q}/\mathbb{Z} \tag*{(137)} A 10 = A − 32 = Z /3 , (( − , − )) : A − 32 × A 10 ⟶ Q / Z (137)
as a perfect pairing, with the Lie-framed Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) image generating A 10 A_{10} A 10 [3 , 21 ] . The nonzero value (134) therefore implies that x x x has exact order three. If the additional u V u_V u V in Hypothesis 10.1 is supplied, restriction to the basepoint places i ∗ u V i^*u_V i ∗ u V in A − 32 A_{-32} A − 32 . Restricted Tate detection gives it the same pairing with the generator; perfectness then implies
i ∗ u V = x D − x E = x . (138) i^*u_V = x_D - x_E = x. \tag*{(138)} i ∗ u V = x D − x E = x . (138)
Here i ∗ i^* i ∗ denotes restriction along p t → B K \mathrm{pt} \to BK pt → B K ; the map i i i in (130) is the induced cofiber map. This is the conditional conclusion of Li–Sun [7 Corollary 10.7] .
The conditional branch thus gives the calibrated third-root response a precise role as a torsion detector. It tests supplied integral assignments using a known perfect pairing, without constructing those assignments or proving an equivalence of supersymmetric quantum field theories with TMF. Its output is compatible with the published heterotic scalar comparison, while its premises remain the separate integral lifts, real characters and selected differential filling specified here.
Consequences for global background dependence
On the ten-manifold M = S ( V ) M = S(V) M = S ( V ) , the normalized Green–Schwarz response depends on the complete background, including its integral differential String datum. The local anomaly polynomial fixes the curvature dependence, while the closed value O ( M ) = exp ( 2 π i ϕ ( M ) ) \mathcal{O}(M) = \exp(2\pi i\phi(M)) O ( M ) = exp ( 2 π i ϕ ( M )) also uses the real fermion normalization, Spin lifts, boundary frames and differential identifications described in § 2 . The sphere-bundle formula derived by Li–Sun [7 ] expresses this dependence through a relative characteristic pairing on the base filling. It makes the integral datum visible in the calculation and determines which changes of that datum the response can distinguish.
In the base-filling notation of § 5 , W W W is a spin eight-manifold filling the seven-dimensional base B B B . The forms c W c_W c W and λ W \lambda_W λ W are the extended gauge and tangent characteristic forms, and h h h is the three-form component of the boundary trivialization. The disk construction and the twelve-dimensional Atiyah–Patodi–Singer (APS) calculation give
ϕ ( S ( V ) ) = 1 3 [ ∫ W c W ( λ W + c W ) − ∫ B h c ] = K 3 ( m o d Z ) . (139) \phi(S(V)) = \frac{1}{3}\left[\int_W c_W(\lambda_W+c_W)-\int_B hc\right] = \frac{K}{3}\pmod{\mathbb{Z}}. \tag*{(139)} ϕ ( S ( V )) = 3 1 [ ∫ W c W ( λ W + c W ) − ∫ B h c ] = 3 K ( mod Z ) . (139)
The local contribution a N = ∫ B h c / 3 a_N = \int_B hc/3 a N = ∫ B h c /3 is generally nonzero. Let z z z be the integral relative class fixed by the boundary nullhomotopy. Combining the disk and closed eta contributions with the sign − a N − μ -a_N-\mu − a N − μ evaluates the expression in brackets as the integral pairing ⟨ c W ⌣ z , [ W , B ] ⟩ \langle c_W\smile z,[W,B]\rangle ⟨ c W ⌣ z , [ W , B ]⟩ . Finite tangent-connection transgressions cancel between the two contributions when the same differential datum and its boundary identifications are transported. The even real APS index removes the half-index ambiguity, with the kernel terms retained. Changing the base filling changes K K K by a multiple of 24, so the response is unchanged. Thus the auxiliary spin/gauge twelve-manifold evaluates the spectral contribution while the boundary nullhomotopy supplies the relative correction. The APS filling itself admits no H H H -trivialization, since X 4 X_4 X 4 has integral 2 on its four-sphere fibre. The phase retains only K K K modulo three; it does not recover the complete background from that residue.
The relation to the Crowley–Goette invariant identifies this retained information with an established secondary invariant on exactly the domain H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 H^3(B;\mathbb{Q})=H^4(B;\mathbb{Q})=0 H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 . There the rational relative lift is unique, and the same product gives t B ( V ) = K / 24 t_B(V)=K/24 t B ( V ) = K /24 , hence ϕ = δ t B ( V ) \phi=\delta t_B(V) ϕ = δ t B ( V ) [7 , 9 ] . Simple connectedness plays no role in this comparison. Outside that domain, the relative formula continues to evaluate the scalar and makes its dependence on the integral trivialization explicit. It therefore describes more backgrounds than the comparison with t B ( V ) t_B(V) t B ( V ) alone.
The examples distinguish calibration from background dependence. For the Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) example, the filling built from the actual quaternionic line gives K = − 2 K=-2 K = − 2 and recovers the known response e 2 π i / 3 e^{2\pi i/3} e 2 π i /3 after the specified Lie-frame and connection transport [3 , 7 ] . Inversion on Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) pulls the right Lie framing back to the left one; the subsequent differential String cylinders retain the endpoint identifications and trivial boundary gauges. These identifications equate the scalar with that of the special right-framed datum of period + 1 +1 + 1 ; the general outward fibre period remains − 2 -2 − 2 . In the fixed-ten-manifold example of § 7 , changes pulled back from the base leave this latter period unchanged and give φ n = 2 n / 3 \varphi_n=2n/3 φ n = 2 n /3 modulo Z \mathbb{Z} Z . For n = 0 , 1 , 2 n=0,1,2 n = 0 , 1 , 2 the phases are distinct, while the integral String classes remain distinct even when n n n changes by three and the phase repeats. These response values concern different background data on one geometry, rather than a count of vacua. The example lies outside the Crowley–Goette domain and exhibits the information a fibre period fails to specify.
The reduced six-dimensional virtual fermion system returns the same closed response on backgrounds trivializing λ + c \lambda+c λ + c . Its primitive order-three value makes 3 ∣ r 3\mid r 3 ∣ r the exact condition for r r r copies, with r r r a nonnegative integer, to have trivial closed response on every permitted background. The Lee–Tachikawa primitive normalization explains this divisibility; the matching integral lattice factorization uses the Monnier–Moore and Monnier–Moore–Park conventions [10 , 11 , 12 ] . With primitive c c c , integral gauge coefficient and characteristic gravitational coefficient, that factorization exists precisely for multiples of three. The density comparison preserves the different background trivializations: the reduced response uses λ + c \lambda+c λ + c , whereas Lee–Tachikawa trivialize c c c . The lattice identity answers the universal coefficient-factorization problem in § 8 ; it supplies neither the quantum tensor fields nor the remaining constraints of an interacting supergravity theory.
The published ordered heterotic comparison matches actual closed real fermion operators within its fermion and Green–Schwarz section prescription, including the prescribed real gluing and exclusion of an independent invertible sector [3 ] . § 9 explains how those inputs relate its scalar to the response computed here. Equality of the response is obtained with that prescription; the scalar alone does not identify complete theories.
The integral application to topological modular forms and vertex operator superalgebras has the separate premises of Hypothesis 10.1 : compatible integral lifts and real equivariant characters with twist t t w = − 2 c t_{\mathrm{tw}}=-2c t tw = − 2 c . For the selected left Lie-framed Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) datum and its zero-counterterm filling, the relative pairing in § 10 is 1 / 3 1/3 1/3 . The known perfect torsion pairing then makes the difference of the restricted classes of the two rank-sixteen lattice inputs an element of exact order three. This detects the class under the stated lift hypotheses; it does not construct those lifts from their characters. The role of the calibrated response in this branch is consequently a test of supplied integral assignments, with the selected differential filling retained as part of the test.
Acknowledgments
We thank Yuji Tachikawa for helpful correspondence on the surrounding TMF program and generously providing his lecture notes. The initial exploration, source synthesis, and drafting of this manuscript were assisted by OpenAI’s GPT-5.6, GPT-6 and GPT-6.1 under the authors’ direction. The authors take responsibility for the mathematical claims and the physical exposition.
Differential trivializations and relative products
The boundary term ∫ B h ∧ c \int_B h \wedge c ∫ B h ∧ c in the response is tied to an integral nullhomotopy, rather than an arbitrary solution of d h = λ + c dh=\lambda+c d h = λ + c . This appendix explains the cochain information that makes that distinction precise. It supports the background definition in § 2 , the induced String datum in § 3 , and the integer pairing in § 5 . We use the differential-cocycle model and product of Redden [15 §§3.1–3.2, 3.4] . For a quaternionic line V V V over a closed spin seven-manifold B B B , with underlying complex rank-two bundle U U U , these constructions give the differential and relative data of Li and Sun’s sphere-bundle response [7 §§2–3, 5] . The conventions are λ = p 1 / 2 \lambda=p_1/2 λ = p 1 /2 , c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) , and the quaternionic real orientation ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) , for which λ ( V R ) = − c \lambda(V_{\mathbb{R}})=-c λ ( V R ) = − c . No rational-cohomology vanishing is needed below.
Cocycles, trivializations, and their differences
Let I I I send a differential form to the real singular cochain obtained by integration. Write δ \delta δ for the singular-cochain coboundary and d d d for the exterior derivative on forms; a check accent denotes a differential cocycle or its trivialization. A degree-four differential cocycle on a manifold L L L has components
X ˇ = ( x Z , x R , X ) , δ x Z = 0 , δ x R = I X − x Z , d X = 0. \check{X}=(x_{\mathbb{Z}},x_{\mathbb{R}},X),\qquad\delta x_{\mathbb{Z}}=0,\qquad\delta x_{\mathbb{R}}=IX-x_{\mathbb{Z}},\qquad dX=0. X ˇ = ( x Z , x R , X ) , δ x Z = 0 , δ x R = I X − x Z , d X = 0.
Here x Z x_{\mathbb{Z}} x Z is an integral four-cochain, x R x_{\mathbb{R}} x R is a real three-cochain, and X X X is its curvature four-form. Integral cochains are included in real cochains where these expressions are added. The characteristic class is [ x Z ] [x_{\mathbb{Z}}] [ x Z ] ; the real-cochain relation records its agreement with the de Rham class of X X X . It does not identify the whole differential cocycle with either component alone.
A geometric trivialization is a triple s ˇ = ( b , k , H ) \check{s}=(b,k,H) s ˇ = ( b , k , H ) satisfying
δ b = x Z , I H − b − δ k = x R , d H = X . (140) \delta b=x_{\mathbb{Z},\qquad IH-b-\delta k=x_{\mathbb{R}},\qquad dH=X.} \tag*{(140)} δ b = x Z , I H − b − δ k = x R , d H = X . (140)
The degrees of b b b , k k k , H H H are respectively three, two, and three, with b b b integral and k k k real. These are Redden’s trivialization equations [15 Propositions 3.11–3.12] . The triple belongs to the differential cochain complex whose form components begin in degree three; the complex defining degree-four differential characters would have no three-form component in degree three. This distinction is what permits nonzero curvature X X X : a geometric trivialization here refines the integral nullhomotopy and has a connection form H H H whose derivative is X X X . It is not a connection-preserving identification of a nonflat cocycle with the zero differential character.
Existence requires the integral characteristic class to vanish. Vanishing only in real cohomology is weaker, and can leave an integral obstruction. Once a trivialization exists, its isomorphism classes form a torsor for the degree-three differential cohomology group H ^ 3 ( L ; Z ) \widehat{H}^{3}(L;\mathbb{Z}) H 3 ( L ; Z ) , while integral trivializations form a torsor for H 3 ( L ; Z ) H^{3}(L;\mathbb{Z}) H 3 ( L ; Z ) [15 Theorem 3.15] . A torsor means that any two choices have a difference in the indicated group, but there is no preferred zero choice. Forgetting geometric refinement is equivariant for H ^ 3 ( L ; Z ) → H 3 ( L ; Z ) \widehat{H}^{3}(L;\mathbb{Z}) \to H^{3}(L;\mathbb{Z}) H 3 ( L ; Z ) → H 3 ( L ; Z ) . Thus a shift of the integral trivialization is part of the background data; it is not made into an isomorphism by solving the same curvature equation again. Torsion shifts are also retained. A torsion element of H 3 ( L ; Z ) H^{3}(L;\mathbb{Z}) H 3 ( L ; Z ) lifts to a flat degree-three differential character through the Bockstein map for the coefficient sequence Z ↪ R → R / Z \mathbb{Z} \hookrightarrow\mathbb{R} \to\mathbb{R}/\mathbb{Z} Z ↪ R → R / Z . It can therefore change the integral choice without changing H H H [15 Proposition 2.2, equations (2.4)–(2.5)] .
Even within a fixed integral choice, the form is constrained by the middle equation of (A.1). On a closed three-cycle σ \sigma σ it gives
∫ σ H = x R ( σ ) + b ( σ ) , b ( σ ) ∈ Z . \int_{\sigma} H = x_{\mathbb{R}}(\sigma) + b(\sigma), \qquad b(\sigma) \in\mathbb{Z}. ∫ σ H = x R ( σ ) + b ( σ ) , b ( σ ) ∈ Z .
The real two-cochain drops out because δ k ( σ ) = 0 \delta k(\sigma) = 0 δ k ( σ ) = 0 . Consequently the form periods have to refine the specified holonomy and integral nullhomotopy. Merely prescribing d H dH d H does not impose this compatibility.
Morphisms distinguish a change of representative from a change of datum. In this model a morphism represented by an integral two-cochain a a a and real one-cochain ℓ \ell ℓ changes a trivialization by
( b , k , H ) ⟼ ( b + δ a , k − a − δ ℓ , H ) . (b, k, H) \longmapsto(b + \delta a, k - a - \delta\ell, H). ( b , k , H ) ⟼ ( b + δ a , k − a − δ ℓ , H ) .
This preserves all three equations. Likewise a specified morphism between representing degree-four cocycles induces an identification of their trivialization categories. That identification depends on the morphism, not just on the equality of differential-cohomology classes [15 Proposition 3.17] . Boundary gluing therefore requires the actual identifications used in the construction.
Extending a prescribed boundary refinement
The torsor structure also explains the geometric extension used in closed cancellation in § 2 . Here an H \mathcal{H} H -structure consists of the Spin and gauge bundles with an integral nullhomotopy of λ T − λ E − λ E ′ \lambda_T - \lambda_E - \lambda_{E'} λ T − λ E − λ E ′ , where the subscripts denote the tangent bundle and the two gauge bundles of § 2 . Suppose the integral H \mathcal{H} H -structure of a closed eleven-manifold C C C has already been extended over a twelve-dimensional bordism Q H Q_{\mathcal{H}} Q H , including its boundary identification. Extend the metric and gauge connections with product collars. Surjectivity of the geometric-to-integral torsor map gives a geometric refinement of that prescribed bulk integral trivialization [15 Theorem 3.15] .
Its boundary value need not initially equal the prescribed geometric refinement on C C C . Their difference has zero characteristic class, because the integral boundary identification was already fixed. The exact sequence
Ω 2 ( C ) / Ω Z 2 ( C ) ⟶ H ^ 3 ( C ; Z ) ⟶ H 3 ( C ; Z ) ⟶ 0 (141) \Omega^{2}(C)/\Omega_{\mathbb{Z}}^{2}(C) \longrightarrow\widehat{H}^{3}(C;\mathbb{Z}) \longrightarrow H^{3}(C;\mathbb{Z}) \longrightarrow0 \tag*{(141)} Ω 2 ( C ) / Ω Z 2 ( C ) ⟶ H 3 ( C ; Z ) ⟶ H 3 ( C ; Z ) ⟶ 0 (141)
represents this difference by a global two-form β \beta β ; Ω Z 2 \Omega_{\mathbb{Z}}^{2} Ω Z 2 denotes closed two-forms with integral periods. The character represented by ( 0 , I β , d β ) (0,\mathcal{I}\beta,d\beta) ( 0 , I β , d β ) has zero integral component and curvature d β d\beta d β . Extend β \beta β using a collar cutoff, and act on the bulk refinement by this character. Its form component adds d β d\beta d β on the boundary, while its real-cochain component supplies the matching change in the middle trivialization equation. This corrects the boundary value without changing the integral class. The specified boundary isomorphism and its cochain representative extend along the collar, so the trivialization equations hold with the prescribed boundary identification [7 §2, Lemma “Extension of the prescribed refinement”] .
The integral extension comes first. No surjectivity of H 3 ( Q H ; Z ) → H 3 ( C ; Z ) H^3(Q_{\mathcal H};\mathbb{Z}) \to H^3(C;\mathbb{Z}) H 3 ( Q H ; Z ) → H 3 ( C ; Z ) is asserted or required. Nor does this argument turn the spin/gauge twelve-filling Q Q Q of § 4 into an H \mathcal H H -filling: those are distinct constructions with distinct premises.
Characteristic morphisms and the tautological boundary frame
For the disk N = D ( V ) N=D(V) N = D ( V ) and Spin bundles F 1 , F 2 F_1,F_2 F 1 , F 2 , the identities
λ ˇ ( V R ) ≃ − c ˇ 2 ( U ) , λ ˇ ( F 1 ⊕ F 2 ) ≃ λ ˇ ( F 1 ) + λ ˇ ( F 2 ) (142) \check{\lambda}(V_{\mathbb{R}}) \simeq-\check{c}_2(U), \qquad\check{\lambda}(F_1\oplus F_2) \simeq\check{\lambda}(F_1)+\check{\lambda}(F_2) \tag*{(142)} λ ˇ ( V R ) ≃ − c ˇ 2 ( U ) , λ ˇ ( F 1 ⊕ F 2 ) ≃ λ ˇ ( F 1 ) + λ ˇ ( F 2 ) (142)
are used with specified characteristic morphisms. Their integral classes and curvature polynomials agree. At the universal level, possible flat differences lie in degree-three R / Z \mathbb{R}/\mathbb{Z} R / Z cohomology of the relevant products of B S p i n B\mathrm{Spin} B Spin and B S p ( 1 ) B\mathrm{Sp}(1) B Sp ( 1 ) , which vanishes. Choose the unit morphism on framed trivial bundles and use its pullbacks and compositions on N N N [7 §3.2] . This normalization specifies the morphisms in the disk construction; it does not assert uniqueness of arbitrary morphisms on the base B B B .
Write π : N → B \pi:N\to B π : N → B for the disk projection and P → B P\to B P → B for the principal right S p ( 1 ) \mathrm{Sp}(1) Sp ( 1 ) bundle defining V V V . The radial gauge connection A A A is trivial in the tautological frame on a boundary collar, while A ∗ A_* A ∗ is the stable tangent connection with product collar geometry. Their connection changes in § 3 have full transgression cochains T ˇ , K ˇ \check{T},\check{K} T ˇ , K ˇ . Write λ ˇ B = λ ˇ ( T B ) \check{\lambda}_B=\check{\lambda}(TB) λ ˇ B = λ ˇ ( T B ) and let c ˇ \check{c} c ˇ use the base gauge connection. With q ˇ = c ˇ 2 ( π ∗ P , A ) \check{q}=\check{c}_2(\pi^*P,A) q ˇ = c ˇ 2 ( π ∗ P , A ) and q ˇ 0 = π ∗ c ˇ \check{q}_0=\pi^*\check{c} q ˇ 0 = π ∗ c ˇ , their equations are
d T ˇ = q ˇ − q ˇ 0 , d K ˇ = λ ˇ ( A ∗ ) − π ∗ ( λ ˇ B − c ˇ ) . (143) d\check{T}=\check{q}-\check{q}_0,\qquad d\check{K}=\check{\lambda}(A_*)-\pi^*(\check{\lambda}_B-\check{c}). \tag*{(143)} d T ˇ = q ˇ − q ˇ 0 , d K ˇ = λ ˇ ( A ∗ ) − π ∗ ( λ ˇ B − c ˇ ) . (143)
The differential here acts on integral, real, and form components. For a quadratic invariant polynomial P \mathcal{P} P , normalized so that its characteristic form is P ( F A , F A ) \mathcal{P}(F_A,F_A) P ( F A , F A ) , a smooth path A t A_t A t on the same actual bundle gives the form transgression
T P ( A 0 , A 1 ) = 2 ∫ 0 1 P ( A ˙ t , F A t ) d t , T_{\mathcal{P}}(A_0,A_1)=2\int_0^1\mathcal{P}(\dot{A}_t,F_{A_t})\,dt, T P ( A 0 , A 1 ) = 2 ∫ 0 1 P ( A ˙ t , F A t ) d t ,
d T P = P ( F A 1 , F A 1 ) − P ( F A 0 , F A 0 ) . dT_{\mathcal{P}}=\mathcal{P}(F_{A_1},F_{A_1})-\mathcal{P}(F_{A_0},F_{A_0}). d T P = P ( F A 1 , F A 1 ) − P ( F A 0 , F A 0 ) .
Here F A t F_{A_t} F A t is the curvature of A t A_t A t . Invariance and the Bianchi identity turn the derivative of the characteristic form into the exterior derivative of the integrand; integration fixes the endpoint sign. This is the finite form change underlying the Chern–Simons variation formula [24 Proposition 3.8] . A connection cylinder, with the chosen endpoint cocycles and their characteristic morphisms, lifts it to a cochain homotopy with the sign in (A.4); the differential Stokes formula retains its two endpoints [15 equation (3.27)] . The form transgression alone does not specify its integral and real components. Those components and the endpoint morphisms are retained in T ~ \widetilde{T} T and K ~ \widetilde{K} K . These homotopies are not isomorphisms between differential characters with unequal curvature.
The base trivialization obeys d h ˇ = λ ˇ B + c ˇ d\check{h}=\check{\lambda}_{B}+\check{c} d h ˇ = λ ˇ B + c ˇ . Adding these equations to its pullback gives the disk refinement already used in the local calculation,
H ~ N = π ∗ h ~ + 2 T ~ + K ~ , d H ~ N = λ ~ ( A ∗ ) + 2 q ~ . \widetilde{H}_{N}=\pi^{*}\widetilde{h}+2\widetilde{T}+\widetilde{K},\qquad d\widetilde{H}_{N}=\widetilde{\lambda}(A_{*})+2\widetilde{q}. H N = π ∗ h + 2 T + K , d H N = λ ( A ∗ ) + 2 q .
The unit vector on M = S ( V ) M=S(V) M = S ( V ) gives the tautological gauge frame τ \tau τ . It supplies a specified morphism γ ~ τ \widetilde{\gamma}_{\tau} γ τ with d γ ~ τ = q ~ ∣ M d\widetilde{\gamma}_{\tau}=\widetilde{q}|_{M} d γ τ = q ∣ M . Its three-form is zero in the normalized framed flat model, but its cochain components still identify that gauge datum with the trivial one. Using the product Spin identification to split off the outward normal, the induced String trivialization is
s ~ M = H ~ N ∣ M − 2 γ ~ τ , d s ~ M = λ ~ ( T M ) . (144) \widetilde{s}_{M}=\widetilde{H}_{N}|_{M}-2\widetilde{\gamma}_{\tau},\qquad d\widetilde{s}_{M}=\widetilde{\lambda}(TM). \tag*{(144)} s M = H N ∣ M − 2 γ τ , d s M = λ ( T M ) . (144)
The factor two removes the two quaternionic gauge summands with their chosen frames. The resulting equation includes their identifications, even though the boundary gauge curvature vanishes. Discarding γ ~ τ \widetilde{\gamma}_{\tau} γ τ because its three-form vanishes would discard part of the induced background.
The relative cone and the integral product
Let W W W be the spin eight-dimensional filling of § 5 , with ∂ W = B \partial W=B ∂ W = B and actual quaternionic extension V ~ \widetilde{V} V . Give B B B the outward boundary orientation. Put λ W = λ ( T W ) \lambda_{W}=\lambda(TW) λ W = λ ( T W ) and c W = c 2 ( U ~ ) c_{W}=c_{2}(\widetilde{U}) c W = c 2 ( U ) , where U ~ \widetilde{U} U is the underlying complex rank-two bundle, and X = λ W + c W X=\lambda_{W}+c_{W} X = λ W + c W , using the specified extending connections. Write its bulk differential cocycle as X ~ = ( x Z , x R , X ) \widetilde{X}=(x_{\mathbb{Z}},x_{\mathbb{R}},X) X = ( x Z , x R , X ) .
For i : B ↪ W i:B\hookrightarrow W i : B ↪ W , the relative cochain complex consists of pairs with differential
d r e l ( u , v ) = ( d u , i ∗ u − d v ) . (145) d_{\mathrm{rel}}(u,v)=(du,i^{*}u-dv). \tag*{(145)} d rel ( u , v ) = ( d u , i ∗ u − d v ) . (145)
The symbols d d d in this formula mean δ \delta δ for singular cochains and the exterior derivative for forms. The integral boundary nullhomotopy b b b makes ( x Z , b ) (x_{\mathbb{Z}},b) ( x Z , b ) a relative cocycle. Its class is the specified z ∈ H 4 ( W , B ; Z ) z\in H^{4}(W,B;\mathbb{Z}) z ∈ H 4 ( W , B ; Z ) , with j z = λ W + c W jz=\lambda_{W}+c_{W} j z = λ W + c W ; here j j j is the map to H 4 ( W ; Z ) H^{4}(W;\mathbb{Z}) H 4 ( W ; Z ) that forgets the boundary trivialization.
The boundary refinement ( b , k , h ) (b,k,h) ( b , k , h ) obeys
δ b = i ∗ x Z , I h − b − δ k = i ∗ x R , d h = i ∗ X . \delta b=i^{*}x_{\mathbb{Z}},\qquad Ih-b-\delta k=i^{*}x_{\mathbb{R}},\qquad dh=i^{*}X. δ b = i ∗ x Z , I h − b − δ k = i ∗ x R , d h = i ∗ X .
Together with the bulk cocycle equation these imply
( I X , I h ) − ( x Z , b ) = d r e l ( x R , − k ) = ( δ x R , i ∗ x R + δ k ) . (146) (IX,Ih)-(x_{\mathbb{Z}},b)=d_{\mathrm{rel}}(x_{\mathbb{R}},-k)
=(\delta x_{\mathbb{R}},i^{*}x_{\mathbb{R}}+\delta k). \tag*{(146)} ( I X , I h ) − ( x Z , b ) = d rel ( x R , − k ) = ( δ x R , i ∗ x R + δ k ) . (146)
The minus sign on k k k is forced by the second component of the cone differential. Thus ( X , h ) (X,h) ( X , h ) is a relative de Rham representative of the real image of this particular integral class, rather than a real lift chosen independently of b b b .
Multiplication must respect the comparison between forms and singular cochains. Write the differential cocycle for c W c_W c W as c ˇ W = ( c Z , c R , c W ) \check{c}_W=(c_Z,c_R,c_W) c ˇ W = ( c Z , c R , c W ) . In the degree-four product used here, Redden’s formula includes a real-cochain correction
X ~ ⋅ c ˇ W = ( x Z ⌣ c Z , x Z ⌣ c R + x R ⌣ I c W + B ( X , c W ) , X ∧ c W ) . (147) \widetilde{X}\cdot\check{c}_W=\left(x_Z\smile c_Z,\ x_Z\smile c_R+x_R\smile I c_W+\mathcal{B}(X,c_W),\ X\wedge c_W\right). \tag*{(147)} X ⋅ c ˇ W = ( x Z ⌣ c Z , x Z ⌣ c R + x R ⌣ I c W + B ( X , c W ) , X ∧ c W ) . (147)
Here B \mathcal{B} B is the natural chain homotopy comparing cup products of integrated forms with integration of wedge products [15 §3.4, equation (3.21)] . Integration of forms is not a strict equality between these two products at the cochain level. The correction makes the differential satisfy the graded product rule.
In particular, put c Z , B = i ∗ c Z c_{Z,B}=i^*c_Z c Z , B = i ∗ c Z and c R , B = i ∗ c R c_{R,B}=i^*c_R c R , B = i ∗ c R . The boundary product has all three components
( b , k , h ) ⋅ c ˇ W ∣ B = ( b ⌣ c Z , B , − b ⌣ c R , B + k ⌣ I c + B ( h , c ) , h ∧ c ) . (b,k,h)\cdot\check{c}_W|_B=\left(b\smile c_{Z,B},\ -b\smile c_{R,B}+k\smile I c+\mathcal{B}(h,c),\ h\wedge c\right). ( b , k , h ) ⋅ c ˇ W ∣ B = ( b ⌣ c Z , B , − b ⌣ c R , B + k ⌣ I c + B ( h , c ) , h ∧ c ) .
The minus sign in its real component comes from the odd degree of the trivialization. Applying the graded product rule and d c ˇ W = 0 d\check{c}_W=0 d c ˇ W = 0 shows that its differential is the boundary restriction of (A.8); this is Redden’s compatibility of trivializations with products [15 Proposition 3.23] . The real component therefore participates in the comparison between the integral product and its form representative. After passage to relative cohomology, this supplies the product of z z z with the absolute integral class c 2 ( U ~ ) c_2(\widetilde{U}) c 2 ( U ) .
Relative integration uses the same boundary sign as (A.6): ( α , β ) (\alpha,\beta) ( α , β ) pairs with the relative fundamental class [ W , B ] [W,B] [ W , B ] as ∫ W α − ∫ B β \int_W\alpha-\int_B\beta ∫ W α − ∫ B β . Indeed a relative exact pair ( d u , i ∗ u − d v ) (du,i^*u-dv) ( d u , i ∗ u − d v ) has zero integral by Stokes and the fact that B B B is closed. Applying this convention to the product gives
K = ⟨ c 2 ( U ~ ) ⌣ z , [ W , B ] ⟩ = ∫ W c W ( λ W + c W ) − ∫ B h ∧ c ∈ Z . (148) K=\left\langle c_2(\widetilde{U})\smile z,[W,B]\right\rangle=\int_W c_W(\lambda_W+c_W)-\int_B h\wedge c\in\mathbb{Z}. \tag*{(148)} K = ⟨ c 2 ( U ) ⌣ z , [ W , B ] ⟩ = ∫ W c W ( λ W + c W ) − ∫ B h ∧ c ∈ Z . (148)
The displayed integrality comes from the integral relative cup-product pairing. The form identity computes its real image; it does not supply integrality on its own. No extension of h ˇ \check{h} h ˇ as a trivialization over W W W is required.
If one retained only d h = λ + c dh=\lambda+c d h = λ + c , the pair ( X , h ) (X,h) ( X , h ) would be closed in the relative form complex, but its identification with ( x Z , b ) (x_Z,b) ( x Z , b ) in (A.7) would be missing. Replacing h h h by h + ρ h+\rho h + ρ for a closed real three-form preserves that curvature equation and changes the displayed relative integral by − ∫ B ρ ∧ c -\int_B\rho\wedge c − ∫ B ρ ∧ c . There is no integral quantization condition on such an arbitrary real change. In the actual background it is the full trivialization equations and the specified morphisms that control which changes are allowed.
Equation (A.9) is the integer entering (64) and the response ϕ ( S ( V ) , s ˇ M ) = K / 3 \phi(S(V),\check{s}_M)=K/3 ϕ ( S ( V ) , s ˇ M ) = K /3 in (65) . Its boundary subtraction contains exactly the h ∧ c h\wedge c h ∧ c term whose local contribution is a V = ∫ B h ∧ c / 3 a_V=\int_B h\wedge c/3 a V = ∫ B h ∧ c /3 in § 3 . This returns the cochain construction to the scalar with its fixed sign ϕ = − a − μ \phi=-a-\mu ϕ = − a − μ , where a = a V a=a_V a = a V and μ \mu μ is the closed real eta contribution defined in § 2 . The rational lift and factor 1 / 24 1/24 1/24 needed for the Crowley–Goette comparison are separate steps in § 5 ; they do not replace the integral datum used here.
Real eta invariants and Pfaffian conventions
The real fermion coefficient accounts for the factor 1 / 2 1/2 1/2 in the anomaly polynomial. The sign of the closed eta phase additionally depends on the chirality and the anomaly line, while first-slot conjugation in the Hermitian reference pairing determines the local sign in ϕ = − a − μ \phi=-a-\mu ϕ = − a − μ . This appendix supplies the character expansion and Atiyah–Patodi–Singer (APS) spectral conventions used in § 2 and § 4 , then translates them into the closed-operator comparison of § 9 . We retain actual real bundles, their Spin lifts and connections, product collars, and kernel terms. The full real anomaly polynomial, real APS period and closed fermion-phase comparison follow the mathematical companion [7 §2, §9] . The real boundary-state construction below fixes its spectral-cut transitions, graded duality and actual seam maps, thereby supplying the contraction used by the normalized section and reference pairing.
The vector and half-spin characters
For a rank-sixteen real Spin bundle E E E , write p = p 1 ( E ) p=p_1(E) p = p 1 ( E ) , q = p 2 ( E ) q=p_2(E) q = p 2 ( E ) and r = p 3 ( E ) r=p_3(E) r = p 3 ( E ) , and use primes for E ′ E' E ′ . If x 1 , … , x 8 x_1,\ldots,x_8 x 1 , … , x 8 are its formal Pontryagin roots, the vector character is ∑ i ( e x i + e − x i ) \sum_i(e^{x_i}+e^{-x_i}) ∑ i ( e x i + e − x i ) . The half-spin character is 128 ∏ i cosh ( x i / 2 ) 128\prod_i\cosh(x_i/2) 128 ∏ i cosh ( x i /2 ) plus a chirality-dependent product of hyperbolic sines. Each x i x_i x i has form degree two, so the latter product first contributes in degree sixteen. Consequently the choice of half-spin chirality does not alter the following degree-twelve expansion, although its actual representation is fixed by the gauge embedding:
ch ( E C ) = 16 + p + p 2 − 2 q 12 + p 3 − 3 p q + 3 r 360 , (149) \operatorname{ch}(E_{\mathbb{C}})=16+p+\frac{p^2-2q}{12}+\frac{p^3-3pq+3r}{360}, \tag*{(149)} ch ( E C ) = 16 + p + 12 p 2 − 2 q + 360 p 3 − 3 pq + 3 r , (149)
ch ( S R + ( E ) ) = 128 + 16 p + p 2 + 4 q 3 + p 3 360 + p q 30 + 2 r 15 . (150) \operatorname{ch}(S_{\mathbb{R}}^+(E))=128+16p+\frac{p^2+4q}{3}+\frac{p^3}{360}+\frac{pq}{30}+\frac{2r}{15}. \tag*{(150)} ch ( S R + ( E )) = 128 + 16 p + 3 p 2 + 4 q + 360 p 3 + 30 pq + 15 2 r . (150)
Here and below character equalities are truncated through degree twelve. The roots use the characteristic-class normalization of § 2 ; in Tachikawa–Zhang’s curvature notation, 1 / ( 2 π ) 1/(2\pi) 1/ ( 2 π ) is absorbed into F F F and ch = tr e i F \operatorname{ch}=\operatorname{tr}e^{iF} ch = tr e i F [3 §2, §4.3] .
Multiply the two vector characters and subtract the two half-spin characters, as required by Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) \Delta_{\mathbb{R}}=E\otimes_{\mathbb{R}}E'-S_{\mathbb{R}}^+(E)-S_{\mathbb{R}}^+(E') Δ R = E ⊗ R E ′ − S R + ( E ) − S R + ( E ′ ) . The degree-zero cancellation is 256 − 128 − 128 = 0 256-128-128=0 256 − 128 − 128 = 0 , and the degree-four cancellation is 16 ( p + p ′ ) − 16 p − 16 p ′ = 0 16(p+p')-16p-16p'=0 16 ( p + p ′ ) − 16 p − 16 p ′ = 0 . In degree eight the remaining terms are
ch 8 ( Δ C ) = p 2 + p p ′ + p ′ 2 − 4 q − 4 q ′ = 4 S , S = n 2 + n n ′ + n ′ 2 − q − q ′ , \operatorname{ch}_8(\Delta_{\mathbb{C}})=p^2+pp'+p'^2-4q-4q'=4S,\qquad S=n^2+nn'+n'^2-q-q', ch 8 ( Δ C ) = p 2 + p p ′ + p ′2 − 4 q − 4 q ′ = 4 S , S = n 2 + n n ′ + n ′2 − q − q ′ ,
where n = p / 2 n=p/2 n = p /2 and n ′ = p ′ / 2 n'=p'/2 n ′ = p ′ /2 . In degree twelve the r , r ′ r,r' r , r ′ terms cancel, while the other terms give
ch 12 ( Δ C ) = p 3 + p ′ 3 24 + p 2 p ′ + p p ′ 2 12 − ( p + p ′ ) ( q + q ′ ) 6 = n + n ′ 3 S . (151) \begin{aligned}
\operatorname{ch}_{12}(\Delta_{\mathbb{C}})
&=\frac{p^3+p'^3}{24}+\frac{p^2p'+pp'^2}{12}-\frac{(p+p')(q+q')}{6}\\
&=\frac{n+n'}{3}S. \tag*{(151)}
\end{aligned} ch 12 ( Δ C ) = 24 p 3 + p ′3 + 12 p 2 p ′ + p p ′2 − 6 ( p + p ′ ) ( q + q ′ ) = 3 n + n ′ S . (151)
Because the character has no degree-zero or degree-four term, only A ^ 4 = − p 1 ( T ) / 24 = − λ ( T ) / 12 \widehat{A}_4=-p_1(T)/24=-\lambda(T)/12 A 4 = − p 1 ( T ) /24 = − λ ( T ) /12 contributes alongside ch 12 \operatorname{ch}_{12} ch 12 . Thus
1 2 [ A ^ ( T ) ch ( Δ C ) ] 12 = 1 2 ( n + n ′ − λ ( T ) 3 S ) = X 4 Y 8 , (152) \frac{1}{2}\left[\widehat{A}(T)\operatorname{ch}(\Delta_{\mathbb{C}})\right]_{12}=\frac{1}{2}\left(\frac{n+n'-\lambda(T)}{3}S\right)=X_4Y_8, \tag*{(152)} 2 1 [ A ( T ) ch ( Δ C ) ] 12 = 2 1 ( 3 n + n ′ − λ ( T ) S ) = X 4 Y 8 , (152)
with X 4 = λ ( T ) − n − n ′ X_4=\lambda(T)-n-n' X 4 = λ ( T ) − n − n ′ and Y 8 = S / 6 Y_8=S/6 Y 8 = S /6 . Both gauge p 2 p_2 p 2 terms survive. In particular, replacing Y 8 Y_8 Y 8 by a polynomial involving only n , n ′ n,n' n , n ′ would change the coefficient system being evaluated.
For the sphere-bundle gauge E = 2 V R ⊕ R 8 E=2V_{\mathbb{R}}\oplus\mathbb{R}^8 E = 2 V R ⊕ R 8 , let c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) for the underlying complex rank-two bundle U U U of V V V . Since p ( V R ) = 1 − 2 c + c 2 p(V_{\mathbb{R}})=1-2c+c^2 p ( V R ) = 1 − 2 c + c 2 , squaring this total Pontryagin class gives n = − 2 c n=-2c n = − 2 c and p 2 ( E ) = 6 c 2 p_2(E)=6c^2 p 2 ( E ) = 6 c 2 . With E ′ E' E ′ trivial, Y 8 = − c 2 / 3 Y_8=-c^2/3 Y 8 = − c 2 /3 . This is the specialization used on the disk and its closed double. A single copy V R ⊕ R 12 V_{\mathbb{R}}\oplus\mathbb{R}^{12} V R ⊕ R 12 instead has n = − c n=-c n = − c , p 2 = c 2 p_2=c^2 p 2 = c 2 , and Y 8 = 0 Y_8=0 Y 8 = 0 ; it cannot supply the same density. This distinction is also explained by the mathematical companion’s comparison with the published specialization [3 §4.3, equation (77)] .
Real representatives and the APS period
For an actual real twist R R R on a closed spin eleven-manifold Z Z Z , let D Z , R C D_{Z,R_{\mathbb{C}}} D Z , R C be the ordinary self-adjoint Levi–Civita Dirac operator and set
h R = dim C ker D Z , R C , ξ R = η A P S ( D Z , R C ) + h R 2 . (153) h_R=\dim_{\mathbb{C}}\ker D_{Z,R_{\mathbb{C}}},\qquad\xi_R=\frac{\eta_{\mathrm{APS}}(D_{Z,R_{\mathbb{C}}})+h_R}{2}. \tag*{(153)} h R = dim C ker D Z , R C , ξ R = 2 η APS ( D Z , R C ) + h R . (153)
The APS eta invariant excludes zero eigenvalues. For a virtual real coefficient, add or subtract these real-valued representatives, including their kernel dimensions, before defining
μ I ( Z ) = − 1 2 ξ Δ ( Z ) ( m o d Z ) . (154) \mu_I(Z)=-\frac{1}{2}\xi_{\Delta}(Z)\pmod{\mathbb{Z}}. \tag*{(154)} μ I ( Z ) = − 2 1 ξ Δ ( Z ) ( mod Z ) . (154)
This construction does not divide an unspecified element of R / Z \mathbb{R}/\mathbb{Z} R / Z by two. Such a division would leave two possible answers; the real spectral system supplies the refinement that selects the phase.
In dimension 11 ≡ 3 ( m o d 8 ) 11\equiv3\pmod{8} 11 ≡ 3 ( mod 8 ) , a real-twisted Spin operator has a compatible antilinear symmetry whose square is − 1 -1 − 1 and which commutes with the operator. Its eigenspaces, including the kernel, therefore have even complex dimension. Freed–Moore construct the resulting half-eta function in Proposition 1.28 [16 ] ; Witten–Yonekura display the antilinear symmetry in equations (2.47)–(2.51) [17 ] . At a crossing of a nonzero eigenvalue through zero the eta change has the even multiplicity required to preserve the real phase. At the crossing itself the kernel contribution in (B.5) is essential.
Orientation reversal sends the closed operator to − D -D − D while preserving its kernel. Hence
ξ ( − D ) = − ξ ( D ) + h R , μ I ( − Z ) = − μ I ( Z ) ( m o d Z ) , (155) \xi(-D)=-\xi(D)+h_R,\qquad\mu_I(-Z)=-\mu_I(Z)\pmod{\mathbb{Z}}, \tag*{(155)} ξ ( − D ) = − ξ ( D ) + h R , μ I ( − Z ) = − μ I ( Z ) ( mod Z ) , (155)
where the second identity uses even h R h_R h R and extends by virtual subtraction. This is a statement about closed real-twisted operators; boundary-line reversal additionally involves the graded identifications discussed below.
For a compact spin twelve-manifold Q Q Q with the actual real twists extended and product metric and connection near its boundary, twisted APS gives
ξ Δ ( ∂ Q ) = ∫ Q [ A ^ ( T Q ) ch ( Δ C ) ] 12 − Ind A P S D Δ C + . (156) \xi_{\Delta}(\partial Q)=\int_Q[\widehat{A}(TQ)\operatorname{ch}(\Delta_{\mathbb{C}})]_{12}-\operatorname{Ind}_{\mathrm{APS}}D^+_{\Delta_{\mathbb{C}}}. \tag*{(156)} ξ Δ ( ∂ Q ) = ∫ Q [ A ( T Q ) ch ( Δ C ) ] 12 − Ind APS D Δ C + . (156)
The index formula is [8 Theorem (4.2), formula (4.3)] . In dimension 12 ≡ 4 ( m o d 8 ) 12\equiv4\pmod{8} 12 ≡ 4 ( mod 8 ) , the real-twisted chiral Spin representations have parallel quaternionic structures compatible with the Dirac operator and the APS spectral projections, including their zero modes. Both chiral APS kernels have even complex dimension; their difference, and then its virtual subtraction, is even. Multiplication of (B.8) by − 1 / 2 -1/2 − 1/2 therefore yields
μ I ( ∂ Q ) = ∫ Q I 12 ( m o d Z ) . (157) \mu_I(\partial Q)=\int_Q I_{12}\pmod{\mathbb{Z}}. \tag*{(157)} μ I ( ∂ Q ) = ∫ Q I 12 ( mod Z ) . (157)
This evenness is the reason the half-index ambiguity is integral [17 (3.8)–(3.9)] . Ordinary APS requires a spin/gauge filling, not an extension of the differential H \mathcal{H} H trivialization. It thus applies to Q Q Q in § 4 , even though ∫ F X 4 = 2 \int_F X_4=2 ∫ F X 4 = 2 prevents that filling from being an H \mathcal{H} H -filling.
The boundary line and the two operator conventions
Fix positive-index chirality by I ( R ) = 1 2 [ A ^ ch ( R C ) ] 12 I(R)=\frac{1}{2}[\widehat{A}\operatorname{ch}(R_{\mathbb{C}})]_{12} I ( R ) = 2 1 [ A ch ( R C ) ] 12 , and denote its boundary fermion Pfaffian line by P R \mathcal{P}_R P R . Freed–Moore’s real state has phase exp ( + π i ξ R ) \exp(+\pi i\xi_R) exp ( + π i ξ R ) and lies in P R ∗ \mathcal{P}_R^* P R ∗ [16 (3.11)] . The positive fermion functional integral lies in P R \mathcal{P}_R P R [16 (4.33)–(4.34)] . Only this line assignment is used from the latter passage, not its subsequent conjectural M-theory cancellation. In Tachikawa–Zhang’s prescription the fermion functional integral and its anomaly state lie in the same line [3 §4.2] . Accordingly its closed fermion-anomaly phase is the dual of the Freed–Moore state:
χ R ( Z ) = exp ( − π i ξ R ( Z ) ) . (158) \chi_R(Z)=\exp(-\pi i\xi_R(Z)). \tag*{(158)} χ R ( Z ) = exp ( − π i ξ R ( Z )) . (158)
The subscripts FM and WY below distinguish the Freed–Moore and Witten–Yonekura operator conventions, respectively. Witten–Yonekura express inflow as exp ( − i π η W Y / 2 ) \exp(-i\pi\eta_{\mathrm{WY}}/2) exp ( − iπ η WY /2 ) [17 (2.10), (2.52)] . Its apparent minus sign cannot be compared with (B.10) without identifying the operators. Use Freed–Moore’s graded Clifford module M = M 0 ⊕ M 1 \mathcal{M}=\mathcal{M}^0\oplus\mathcal{M}^1 M = M 0 ⊕ M 1 , with geometric Dirac operator D = c ∘ ∇ D=c\circ\nabla D = c ∘ ∇ . In odd dimension both graded spinor summands are copies of the ungraded spinor bundle, tensored with the coefficient R C R_{\mathbb{C}} R C . The operator D D D exchanges the summands. Here c ( v ) c(v) c ( v ) denotes skew Clifford multiplication, distinct from the characteristic class c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) above. Let ω \omega ω be the normalized odd volume action and ω ∂ \omega_{\partial} ω ∂ the boundary volume action. For the outward unit normal ν \nu ν on the product collar, their conventions give
ω = c ( ν ) ω ∂ , ω 2 = 1 , c ( ν ) 2 = ω ∂ 2 = − 1. (159) \omega=c(\nu)\omega_{\partial},\qquad\omega^2=1,\qquad c(\nu)^2=\omega_{\partial}^2=-1. \tag*{(159)} ω = c ( ν ) ω ∂ , ω 2 = 1 , c ( ν ) 2 = ω ∂ 2 = − 1. (159)
The volume action commutes with c ( v ) c(v) c ( v ) . On the even Clifford-module summand M 0 \mathcal{M}^0 M 0 , define the Hermitian Witten–Yonekura matrices by
γ W Y ( v ) = i ω c ( v ) , γ W Y ( ν ) = − i ω ∂ . (160) \gamma_{\mathrm{WY}}(v)=i\omega c(v),\qquad\gamma_{\mathrm{WY}}(\nu)=-i\omega_{\partial}. \tag*{(160)} γ WY ( v ) = iω c ( v ) , γ WY ( ν ) = − i ω ∂ . (160)
They satisfy the positive Clifford relation. Witten–Yonekura’s + 1 +1 + 1 normal mode therefore agrees with Freed–Moore’s + i +i + i boundary subspace. Composing D D D with ω \omega ω returns to M 0 \mathcal{M}^{0} M 0 , giving a self-adjoint operator on one copy of the twisted odd spinors. The two closed odd operators have opposite signs:
d F M = ω D ∣ M 0 , D W Y = i γ W Y a ∇ a = − d F M . (161) d_{\mathrm{FM}}=\left.\omega D\right|_{\mathcal{M}^{0}},\qquad D_{\mathrm{WY}}=i\gamma_{\mathrm{WY}}^{a}\nabla_{a}=-d_{\mathrm{FM}}. \tag*{(161)} d FM = ω D ∣ M 0 , D WY = i γ WY a ∇ a = − d FM . (161)
The modified tangential operator in Witten–Yonekura is also the negative of Freed–Moore’s; boundary grading intertwines these tangential operators without reversing the graded index. Thus agreement of the displayed boundary chirality does not imply equality of the closed eta representatives.
Witten–Yonekura count zero modes with positive sign [17 (2.19) and following paragraph] . For h R = dim C ker d F M h_{R}=\dim_{\mathbb{C}}\ker d_{\mathrm{FM}} h R = dim C ker d FM their convention therefore gives
η W Y = − η A P S ( d F M ) + h R = − 2 ξ R + 2 h R , \eta_{\mathrm{WY}}=-\eta_{\mathrm{APS}}(d_{\mathrm{FM}})+h_{R}=-2\xi_{R}+2h_{R}, η WY = − η APS ( d FM ) + h R = − 2 ξ R + 2 h R ,
exp ( − i π η W Y / 2 ) = exp ( + π i ξ R ) . (162) \exp(-i\pi\eta_{\mathrm{WY}}/2)=\exp(+\pi i\xi_{R}). \tag*{(162)} exp ( − iπ η WY /2 ) = exp ( + π i ξ R ) . (162)
The last equality uses even h R h_{R} h R . Witten–Yonekura inflow is thus the Freed–Moore inverse-line state, whose dual is (B.10). The familiar statement η W Y = 2 ξ \eta_{\mathrm{WY}}=2\xi η WY = 2 ξ holds for Witten–Yonekura’s own odd operator; using it for d F M d_{\mathrm{FM}} d FM would erase the sign in (B.13).
The closed coefficient and finite tangent transport
On the actual common cover G = Spin ( 16 ) × Spin ( 16 ) G=\operatorname{Spin}(16)\times\operatorname{Spin}(16) G = Spin ( 16 ) × Spin ( 16 ) , the real associated bundles decompose as
ad ( Spin ( 32 ) ) ∣ G = Λ 2 E ⊕ ( E ⊗ R E ′ ) ⊕ Λ 2 E ′ , \operatorname{ad}(\operatorname{Spin}(32))|_{G}=\Lambda^{2}E\oplus(E\otimes_{\mathbb{R}}E')\oplus\Lambda^{2}E', ad ( Spin ( 32 )) ∣ G = Λ 2 E ⊕ ( E ⊗ R E ′ ) ⊕ Λ 2 E ′ ,
ad ( E 8 × E 8 ) ∣ G = Λ 2 E ⊕ S R + ( E ) ⊕ Λ 2 E ′ ⊕ S R + ( E ′ ) . (163) \operatorname{ad}(E_{8}\times E_{8})|_{G}=\Lambda^{2}E\oplus S_{\mathbb{R}}^{+}(E)\oplus\Lambda^{2}E'\oplus S_{\mathbb{R}}^{+}(E'). \tag*{(163)} ad ( E 8 × E 8 ) ∣ G = Λ 2 E ⊕ S R + ( E ) ⊕ Λ 2 E ′ ⊕ S R + ( E ′ ) . (163)
The first decomposition is the skew-matrix block decomposition. In the second, the 112 roots ± e i ± e j \pm e_i\pm e_j ± e i ± e j and eight Cartan directions give the D 8 D_{8} D 8 adjoint; the other 128 roots with half-integral coordinates and even sign parity give the embedding’s real half-spin representation. Invariant inner products make these equivariant decompositions orthogonal, and equivariance makes them parallel for the induced connections. Consequently they identify actual closed Dirac systems, including holonomy and kernels. Under the fermion prescription of § 9 , the common adjoint, gravitational and ghost factors cancel with identical inverse factors. The remaining coefficient is Δ R \Delta_{R} Δ R up to the common gaugino chirality.
Equation (B.4) gives I ( Δ R ) = − X 4 Y 8 I(\Delta_{R})=-X_{4}Y_{8} I ( Δ R ) = − X 4 Y 8 , while the ordered published difference has + X 4 Y 8 +X_{4}Y_{8} + X 4 Y 8 [3 (74)–(76)] . This nonzero universal polynomial fixes the remaining binary chirality to − Δ R -\Delta_{R} − Δ R after the operator system has been identified. It does not exclude an independent flat anomaly sector; the stated physical prescription excludes such an addition. Direct sums add eta invariants and kernel dimensions, and virtual subtraction inverts their phases, so
I d i f f ( Z ) = χ − Δ R ( Z ) = exp ( + π i ξ Δ ( Z ) ) = exp ( − 2 π i μ I ( Z ) ) . (164) I_{\mathrm{diff}}(Z)=\chi_{-\Delta_{R}}(Z)=\exp(+\pi i\xi_{\Delta}(Z))=\exp(-2\pi i\mu_{I}(Z)). \tag*{(164)} I diff ( Z ) = χ − Δ R ( Z ) = exp ( + π i ξ Δ ( Z )) = exp ( − 2 π i μ I ( Z )) . (164)
This is a comparison of the prescribed closed systems, not an equivalence of arbitrary anomaly theories or their families.
The tangent connection used in the form calculation may be a specified stable metric Spin connection A ∗ A_* A ∗ rather than Levi–Civita. Keep the gauge connections fixed and put
α = CS λ ( A L C , A ∗ ) , d α = λ ( A ∗ ) − λ ( A L C ) , H ∗ = H L C + α . \alpha= \operatorname{CS}_{\lambda}(A_{\mathrm{LC}}, A_*), \qquad\mathrm{d}\alpha= \lambda(A_*) - \lambda(A_{\mathrm{LC}}), \qquad H_* = H_{\mathrm{LC}} + \alpha. α = CS λ ( A LC , A ∗ ) , d α = λ ( A ∗ ) − λ ( A LC ) , H ∗ = H LC + α .
The last equality transports the existing differential trivialization with its integral class and boundary morphism. Since Y 8 Y_8 Y 8 is gauge-only, the full index transgression is α Y 8 \alpha Y_8 α Y 8 . On Z = N ‾ ∪ M W r e f Z = \overline{N} \cup_M W_{\mathrm{ref}} Z = N ∪ M W ref , the collar paths match and Y 8 Y_8 Y 8 vanishes on the trivial-gauge reference. Hence
μ ∗ ( Z ) − μ L C ( Z ) = ∫ Z α Y 8 = − ∫ N α Y 8 ( m o d Z ) , \mu_*(Z) - \mu_{\mathrm{LC}}(Z) = \int_Z \alpha Y_8 = -\int_N \alpha Y_8 \pmod{\mathbb{Z}}, μ ∗ ( Z ) − μ LC ( Z ) = ∫ Z α Y 8 = − ∫ N α Y 8 ( mod Z ) ,
a ∗ − a L C = ∫ N α Y 8 , ( a + μ ) ∗ = ( a + μ ) L C ( m o d Z ) . (165) a_* - a_{\mathrm{LC}} = \int_N \alpha Y_8, \qquad(a+\mu)_* = (a+\mu)_{\mathrm{LC}} \pmod{\mathbb{Z}}. \tag*{(165)} a ∗ − a LC = ∫ N α Y 8 , ( a + μ ) ∗ = ( a + μ ) LC ( mod Z ) . (165)
The stable value is an ordinary spectral invariant plus a finite transgression, not the spectrum of a proposed torsionful operator. Neither term in the last equality is separately invariant. In particular, changing only the eta sign produces − a + μ -a+\mu − a + μ , whereas complex conjugation produces a + μ a+\mu a + μ ; these agree only when 2 a ∈ Z 2a \in\mathbb{Z} 2 a ∈ Z , a condition absent from the general construction.
Real boundary states and graded gluing
We now identify the real boundary state, including its phase and metric, before using it in a contraction. Work first with an honest real coefficient R R R on a compact spin eleven-manifold Y Y Y with product boundary data. Write A = D F M ∂ A = D_{\mathrm{FM}}^{\partial} A = D FM ∂ for the tangential operator and C C C for the parallel quaternionionic structure. This C C C is distinct from normal Clifford multiplication. On the boundary, C C C exchanges the + i +i + i and − i -i − i eigenspaces of ω ∂ \omega_{\partial} ω ∂ ; neither chiral space is itself quaternionic [16 §3.1, (3.2)–(3.7)] .
Allowed domains and real eta coordinates
The product collar has D = c ( ν ) ( ∂ ν − A ) D = c(\nu)(\partial_\nu- A) D = c ( ν ) ( ∂ ν − A ) . Substituting (B.12) into the Witten–Yonekura tangential operator gives D W , W Y = − A D_{\mathrm{W,WY}} = -A D W , WY = − A , as well as D W Y = − d F M D_{\mathrm{WY}} = -d_{\mathrm{FM}} D WY = − d FM . Witten–Yonekura allow positive tangential eigenvalues [17 §2.2, p. 11] ; these are the negative eigenvalues of A A A . Let ψ ± \psi_\pm ψ ± denote the positive and negative chiral parts of the boundary restriction. Writing A + A^+ A + for the map between these chiral spaces, their allowed high modes therefore satisfy precisely Freed–Moore’s boundary condition
ψ − + A + ∣ A ∣ ψ + = 0 on the nonzero high modes , ∣ A ∣ = ( A 2 ) 1 / 2 . \psi_- + \frac{A^+}{|A|}\psi_+ = 0 \quad\text{on the nonzero high modes}, \qquad|A| = (A^2)^{1/2}. ψ − + ∣ A ∣ A + ψ + = 0 on the nonzero high modes , ∣ A ∣ = ( A 2 ) 1/2 .
This identifies the allowed subspace, rather than changing a sign on the basis of a Hamiltonian description of the modes that are projected out [16 (3.10)] .
For the low modes choose an orthonormal positive chiral basis e i e_i e i and the negative basis f i = C e i f_i = Ce_i f i = C e i . Witten–Yonekura’s low-mode condition [17 (2.65), (2.75)] can be written using independent coefficients a i a^i a i , b i b_i b i as
ψ + = ∑ i a i e i , ψ − = − i ∑ i b i f i , b i = ∑ j K i j a j , K T = − K , K † K = 1. \psi_+ = \sum_i a^i e_i, \qquad\psi_- = -i \sum_i b_i f_i, \qquad b_i = \sum_j K_{ij}a^j, \qquad K^T = -K, \qquad K^\dagger K = 1. ψ + = i ∑ a i e i , ψ − = − i i ∑ b i f i , b i = j ∑ K ij a j , K T = − K , K † K = 1.
The coefficients b i b_i b i are independent complex variables before the boundary condition is imposed; they are not complex conjugates of a i a^i a i on a general Euclidean filling. The corresponding Freed–Moore low-mode map is I = i K I=iK I = i K , with graph ψ − = − I ψ + \psi_-=-I\psi_+ ψ − = − I ψ + . It is skew and unitary for the boundary symplectic pairing
⟨ ⟨ x , y ⟩ ⟩ = ∫ ∂ Y ⟨ C x , y ⟩ d v o l ∂ Y , \langle\!\langle x,y\rangle\!\rangle=\int_{\partial Y}\langle Cx,y\rangle\,\mathrm{dvol}_{\partial Y}, ⟨ ⟨ x , y ⟩ ⟩ = ∫ ∂ Y ⟨ C x , y ⟩ dvol ∂ Y ,
where the Hermitian pairing is conjugate-linear in its first slot. This bilinear form pairs the two chiral spaces nondegenerately. The low map exists because the total boundary mod-two index of the real-twisted filling vanishes; separate boundary components need not have even chiral kernel dimension.
In these coordinates C ( u , v ) = ( − v ˉ , u ˉ ) C(u,v)=(-\bar v,\bar u) C ( u , v ) = ( − v ˉ , u ˉ ) and I I ˉ = − 1 I\bar I=-1 I I ˉ = − 1 . Thus C C C preserves v = − I u v=-Iu v = − I u , and it also preserves the high-mode graph. Since it commutes with d F M d_{\mathrm{FM}} d FM , the boundary-conditioned bulk eigenspaces, including its kernel, are quaternionic. Equation (B.14) therefore holds on these identical boundary domains:
τ R ( Y ; a , I ) = exp ( π i ξ R ( Y ; a , I ) ) = exp ( − i π η W Y ( Y ; a , I ) / 2 ) . \tau_{\mathbb{R}}(Y;a,I)=\exp\left(\pi i\xi_{\mathbb{R}}(Y;a,I)\right)=\exp\left(-i\pi\eta_{\mathrm{WY}}(Y;a,I)/2\right). τ R ( Y ; a , I ) = exp ( π i ξ R ( Y ; a , I ) ) = exp ( − iπ η WY ( Y ; a , I ) /2 ) .
Here ( a , I ) (a,I) ( a , I ) specifies the spectral cut and low-mode map defined below, and ξ R ( Y ; a , I ) \xi_{\mathbb{R}}(Y;a,I) ξ R ( Y ; a , I ) includes the kernel term of the boundary-conditioned self-adjoint operator. These are actual real half-eta coordinates. Freed–Moore’s construction assembles them into a unit state in P R ( ∂ Y ) ∗ \mathcal{P}_{\mathbb{R}}(\partial Y)^* P R ( ∂ Y ) ∗ [16 (3.11)–(3.12)] .
Pfaffian frames and spectral cuts
Choose a > 0 a>0 a > 0 outside the spectrum of A 2 A^2 A 2 , and let E a E_a E a be its finite-dimensional + i +i + i chiral space spanned by eigenvectors of A 2 A^2 A 2 with eigenvalue below a a a . The boundary pairing identifies the opposite chiral space with E a ∗ E_a^* E a ∗ , so the Pfaffian line in this chart is P R , a = det E a ∗ \mathcal{P}_{\mathbb{R},a}=\det E_a^* P R , a = det E a ∗ . For b > a b>a b > a also outside the spectrum, write F a , b = E b ⊖ E a F_{a,b}=E_b\ominus E_a F a , b = E b ⊖ E a and regard B = A + ∣ F a , b B=A^+|_{F_{a,b}} B = A + ∣ F a , b as an invertible skew form under that pairing. Skew invertibility gives dim C F a , b = 2 m \dim_{\mathbb{C}}F_{a,b}=2m dim C F a , b = 2 m for an integer m m m ; its polar map and Pfaffian obey
Q a , b = B ( B † B ) − 1 / 2 , q a , b = Pf ( B ) = r a , b Pf ( Q a , b ) , r a , b = det ( B † B ) 1 / 4 > 0. (166) Q_{a,b}=B(B^\dagger B)^{-1/2},\qquad q_{a,b}=\operatorname{Pf}(B)=r_{a,b}\operatorname{Pf}(Q_{a,b}),\qquad r_{a,b}=\det(B^\dagger B)^{1/4}>0. \tag*{(166)} Q a , b = B ( B † B ) − 1/2 , q a , b = Pf ( B ) = r a , b Pf ( Q a , b ) , r a , b = det ( B † B ) 1/4 > 0. (166)
To see the last equality, put B B B in unitary skew normal form. Each two-dimensional block has one positive singular value repeated twice; dividing that block by the singular value leaves its unitary skew phase. Multiplying the block Pfaffians gives the displayed positive norm and retains every complex phase. The fourth root is of a positive determinant, so it selects no phase branch.
The transition from cut a a a to cut b b b is p ↦ p ⊗ q a , b p\mapsto p\otimes q_{a,b} p ↦ p ⊗ q a , b in the prescribed determinant order. Let ρ a > 0 \rho_a>0 ρ a > 0 be the high-spectrum Pfaffian norm factor,
ρ a = det ζ ( A − A + ∣ E a ⊥ ) 1 / 4 , ρ a = r a , b ρ b . \rho_a=\det_{\zeta}\left(A^-A^+|_{E_a^\perp}\right)^{1/4},\qquad\rho_a=r_{a,b}\rho_b. ρ a = ζ det ( A − A + ∣ E a ⊥ ) 1/4 , ρ a = r a , b ρ b .
Here A − = ( A + ) † A^-=(A^+)^\dagger A − = ( A + ) † , and the determinant is zeta regularized on the positive chiral complement. This is the Pfaffian metric correction of [16 (1.5)–(1.10), (1.23)–(1.25)] . For a skew-unitary low map I I I , the unit Pfaffian frame is u a ( I ) = Pf ( I ) / ρ a u_a(I) = \operatorname{Pf}(I)/\rho_a u a ( I ) = Pf ( I ) / ρ a . The scalar τ R ( Y ; a , I ) \tau_R(Y; a, I) τ R ( Y ; a , I ) represents the inverse-line state in the dual unit frame u a ( I ) ∗ u_a(I)^* u a ( I ) ∗ . Stabilization by the newly included block gives
u a ( I ) ⊗ q a , b = u b ( I ⊕ Q a , b ) . u_a(I) \otimes q_{a,b} = u_b(I \oplus Q_{a,b}). u a ( I ) ⊗ q a , b = u b ( I ⊕ Q a , b ) .
Thus the unit frame and the eta coordinate are unchanged when the cut is enlarged without changing the allowed domain. For a different low map I ′ I' I ′ at the larger cut the inverse-line coordinates satisfy
τ R ( Y ; b , I ′ ) τ R ( Y ; a , I ) = Pf ( I ′ ) Pf ( I ) ⊗ Pf ( Q a , b ) . \frac{\tau_R(Y; b, I')}{\tau_R(Y; a, I)} = \frac{\operatorname{Pf}(I')}{\operatorname{Pf}(I) \otimes\operatorname{Pf}(Q_{a,b})}. τ R ( Y ; a , I ) τ R ( Y ; b , I ′ ) = Pf ( I ) ⊗ Pf ( Q a , b ) Pf ( I ′ ) .
The quotient uses the ordered determinant identification above. For three cuts, the block-product identity q a , c = q a , b ⊗ q b , c q_{a,c} = q_{a,b} \otimes q_{b,c} q a , c = q a , b ⊗ q b , c supplies the cocycle.
The phase in this transition is visible in Witten–Yonekura’s nonzero-mode calculation [17 (2.60)–(2.63)] . With ε = ( 0 1 − 1 0 ) \varepsilon= \begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix} ε = ( 0 − 1 1 0 ) , a kinetic block is i λ ε i\lambda\varepsilon iλ ε , λ > 0 \lambda> 0 λ > 0 . Since D W , W Y = − A D_{\mathrm{W,WY}} = -A D W , WY = − A , the corresponding block here is B = − i λ ε B = -i\lambda\varepsilon B = − iλ ε , whose Pfaffian is − i λ -i\lambda − iλ . Its norm is λ \lambda λ and its polar Pfaffian is − i -i − i . The phase is therefore retained even though the regulated fermion magnitude contributes only the positive factor λ \lambda λ .
A basis change e ′ = e U e' = eU e ′ = e U , f ′ = f U ‾ f' = f\overline{U} f ′ = f U that keeps the standard K K K in the new coordinates changes the physical low map to I U = U ‾ I U T I_U = \overline{U} I U^{\mathsf{T}} I U = U I U T . Hence
Pf ( I U ) Pf ( I ) = det ( U ) − 1 . \frac{\operatorname{Pf}(I_U)}{\operatorname{Pf}(I)} = \det(U)^{-1}. Pf ( I ) Pf ( I U ) = det ( U ) − 1 .
This is exactly the half-eta transformation described after [17 (2.86)] ; the ordered zero-mode insertion transforms by det U \det U det U . Two such congruences giving the same skew form differ by a compact symplectic transformation of determinant one. Neither this transition nor the stabilization discards the factor Pf ( i K ) = i dim E a / 2 Pf ( K ) \operatorname{Pf}(iK) = i^{\dim E_a/2}\operatorname{Pf}(K) Pf ( i K ) = i d i m E a /2 Pf ( K ) .
At a kernel crossing use a fixed positive cut containing the crossing modes. Then E a E_a E a has locally constant rank, ρ a \rho_a ρ a stays smooth and positive, and Pf ( I ) \operatorname{Pf}(I) Pf ( I ) is nonzero. There is no division by the low-mode Pfaffian of A + A^+ A + , which can vanish. For the boundary-conditioned bulk operator, a quaternionic eigenvalue pair crossing zero changes ordinary eta by four and reduced ξ \xi ξ by two; exp ( π i ξ ) \exp(\pi i\xi) exp ( π i ξ ) remains continuous, including at zero because the kernel term is retained. The analytic family construction in [16 Proposition 1.28, §3.2] supplies its smooth unit state. Witten–Yonekura’s nonzero APS overlap (2.80) and their discussion after (3.5) give the corresponding zero-mode extension [17 §§2.5, 3.1] .
Ordered duality and Hermitian contraction
For each ordered boundary component M j M_j M j , let p j = dim C ker A j + m o d 2 p_j = \dim_{\mathbb{C}} \ker A_j^+ \bmod2 p j = dim C ker A j + mod 2 be its Pfaffian-line parity. Interchanging two zero-mode blocks of dimensions b r , b s b_r, b_s b r , b s gives ( − 1 ) b r b s (-1)^{b_r b_s} ( − 1 ) b r b s . Reversing the order of the dual factors therefore gives ( − 1 ) ∑ r < s p r p s (-1)^{\sum_{r<s} p_r p_s} ( − 1 ) ∑ r < s p r p s . Two odd components already produce a minus sign even when total parity is even. A change of one component’s Spin identification by − 1 -1 − 1 acts by U = − 1 U = -1 U = − 1 on that kernel and contributes ( − 1 ) p j (-1)^{p_j} ( − 1 ) p j . The gauge identification at the seam is likewise specified, rather than inferred from characteristic classes [17 p. 30 and footnotes 16, 23] . These are the actual real component degrees and seam maps. The complex determinant-line gluing of Dai–Freed is a graded supertrace [25 Proposition 2.15, Theorem 2.20] . Its erratum corrects the odd collar operator to B ∂ X = J − 1 D ∂ X B_{\partial X}=J^{-1}D_{\partial X} B ∂ X = J − 1 D ∂ X , where J J J is normal Clifford multiplication, and corrects the boundary and cylinder formulas [20 (1.2′), (4.8′)] . Its orientation pairing has ( − 1 ) ( k 2 ) (-1)^{\binom{k}{2}} ( − 1 ) ( 2 k ) , with k k k the number of odd-index components, from reversal of the dual tensor order [20 p. 2344] . That ordinary complex-index grade must not replace p j p_j p j : the real structure exchanges the chiral kernels, so their complex index can vanish while p j p_j p j is odd. We use the matched real Witten–Yonekura and Freed–Moore domains and coordinates above, with these ordered duality and seam conventions, rather than a square root of the complex gluing formula.
Choose the same allowed APS domain on the common boundary of two fillings. Let ∣ Y ⟩ |Y\rangle ∣ Y ⟩ and ∣ Y ′ ⟩ |Y'\rangle ∣ Y ′ ⟩ be the corresponding regulated massive-fermion states, proportional to the normalized ground state ∣ Ω ⟩ |\Omega\rangle ∣Ω ⟩ . Witten–Yonekura express their physical overlap [17 (3.3)–(3.4)] as
⟨ Y ′ ∣ Y ⟩ = ⟨ Y ′ ∣ A P S ⟩ ⟨ A P S ∣ Y ⟩ ∣ ⟨ A P S ∣ Ω ⟩ ∣ 2 . \langle Y' \mid Y\rangle= \frac{\langle Y' \mid\mathrm{APS}\rangle\langle\mathrm{APS} \mid Y\rangle}{|\langle\mathrm{APS} \mid\Omega\rangle|^2}. ⟨ Y ′ ∣ Y ⟩ = ∣ ⟨ APS ∣ Ω ⟩ ∣ 2 ⟨ Y ′ ∣ APS ⟩ ⟨ APS ∣ Y ⟩ .
The denominator is positive, and their (2.80) keeps it nonzero with zero modes. The phase is the ratio of the same APS eta coordinates, hence the closed glued phase by their (3.2). The unit-frame identification just proved identifies this ratio with the Hermitian contraction of the real inverse-Pfaffian states. A change of cut or frame multiplies both coordinates by the same unit scalar and cancels in that contraction. The orientation-dual identification uses reversed dual tensor order; retaining the original order instead requires the explicit Koszul factor above. All comparisons use the same actual Spin and gauge seam maps.
Finally take the ordered tensor product of the honest real summand states and their duals for Δ R \Delta_{\mathrm{R}} Δ R . Write τ Δ , L C ( Y ) \tau_{\Delta,\mathrm{LC}}(Y) τ Δ , LC ( Y ) for this unit state. Transport it to the specified stable connection by
τ Δ , ∗ ( Y ) = exp [ − 2 π i ∫ Y T I 12 ( A L C , A ∗ ) ] τ Δ , L C ( Y ) . (167) \tau_{\Delta,*}(Y)=\exp\left[-2\pi i\int_Y T_{I_{12}}(A_{\mathrm{LC}},A_*)\right]\tau_{\Delta,\mathrm{LC}}(Y). \tag*{(167)} τ Δ , ∗ ( Y ) = exp [ − 2 π i ∫ Y T I 12 ( A LC , A ∗ ) ] τ Δ , LC ( Y ) . (167)
The boundary lines are identified along the prescribed connection homotopy, with its endpoint maps. Matched collar paths make this finite factor additive under gluing. On a closed manifold its value is exp ( − 2 π i μ ∗ ) \exp(-2\pi i\mu_*) exp ( − 2 π i μ ∗ ) , whereas the untransported state has value exp ( + π i ξ Δ ) = exp ( − 2 π i μ I ) \exp(+\pi i\xi_{\Delta})=\exp(-2\pi i\mu_I) exp ( + π i ξ Δ ) = exp ( − 2 π i μ I ) . Consequently, for the specified filling and spin reference,
⟨ τ Δ , ∗ ( N ) , τ Δ , ∗ ( W r e f ) ⟩ = exp [ − 2 π i μ ∗ ( N ∪ M W r e f ‾ ) ] . (168) \langle\tau_{\Delta,*}(N),\tau_{\Delta,*}(W_{\mathrm{ref}})\rangle=\exp\left[-2\pi i\mu_*(\overline{N\cup_M W_{\mathrm{ref}}})\right]. \tag*{(168)} ⟨ τ Δ , ∗ ( N ) , τ Δ , ∗ ( W ref )⟩ = exp [ − 2 π i μ ∗ ( N ∪ M W ref ) ] . (168)
The first slot is conjugate-linear, so its filling is orientation reversed. This establishes the real contraction needed in § 2 , where ordered graded evaluation and capped gluing give the required bordism composition, normalized section and reference formula. Combining that section’s local factor with (B.20) gives O ( M ) = exp [ − 2 π i ( a N + μ ∗ ) ] \mathcal{O}(M)=\exp[-2\pi i(a_N+\mu_*)] O ( M ) = exp [ − 2 π i ( a N + μ ∗ )] , with the local integral retained. § 9 uses this contraction for its prescribed scalar; no equivalence of complete anomaly functors is required. The reference is a spin filling and need not extend the differential H \mathcal{H} H structure.
Lie framings and finite transport on Sp(2)
The relative pairing in § 6 gives K = − 2 K=-2 K = − 2 for the quaternionic sphere bundle over S 7 S^7 S 7 . Identifying that answer with the Lie-framed response requires more than the underlying diffeomorphism S ( V ) ≅ Sp ( 2 ) S(V)\cong\operatorname{Sp}(2) S ( V ) ≅ Sp ( 2 ) : the stable tangent frame, its Spin lift, the framed gauge connections and the integral differential String trivialization must agree. The tautological frame on the sphere bundle gives the left Lie framing. A full Spin gauge connection supplies the right-framed positive-unit datum, and inversion relates their complete differential backgrounds [7 §6] . These identifications connect the actual-line evaluation to the known third-root response [3 §3.2, equation (63)] . Write s L , s R s_L,s_R s L , s R for the integral String structures determined by the left and right Lie framings.
Throughout, B = S ( H 2 ) B=S(\mathbb{H}^2) B = S ( H 2 ) has its outward orientation, P = Sp ( 2 ) → B P=\operatorname{Sp}(2)\to B P = Sp ( 2 ) → B sends p p p to p e 2 pe_2 p e 2 , and the right stabilizer is K = Sp ( 1 ) K=\operatorname{Sp}(1) K = Sp ( 1 ) embedded by u ↦ diag ( u , 1 ) u\mapsto\operatorname{diag}(u,1) u ↦ diag ( u , 1 ) . Here e 1 , e 2 e_1,e_2 e 1 , e 2 are the standard coordinate vectors of H 2 \mathbb{H}^2 H 2 . The actual quaternionic line is V y = y ⊥ H V_y=y^{\perp_{\mathbb{H}}} V y = y ⊥ H ; N = D ( V ) N=D(V) N = D ( V ) , with projection π : N → B \pi:N\to B π : N → B , is oriented by base then real quaternionic fibre, and M = ∂ N = S ( V ) M=\partial N=S(V) M = ∂ N = S ( V ) has the outward boundary orientation. The real fibre orientation is ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) , whereas the complex structure is right multiplication by i i i . Thus e ( V R ) = − c 2 ( U ) e(V_{\mathbb{R}})=-c_2(U) e ( V R ) = − c 2 ( U ) , with U U U the corresponding complex rank-two bundle, and λ ( V R ) = − c 2 ( U ) \lambda(V_{\mathbb{R}})=-c_2(U) λ ( V R ) = − c 2 ( U ) with λ = p 1 / 2 \lambda=p_1/2 λ = p 1 /2 . The stabilizer K K K and the relative integer K K K have different roles.
The tautological frame and the integral framing shift
At y ∈ S ( H 2 ) y\in S(\mathbb{H}^2) y ∈ S ( H 2 ) , the decomposition T y B = y ⊥ H ⊕ y Im H T_yB=y^{\perp_{\mathbb{H}}}\oplus y\operatorname{Im}\mathbb{H} T y B = y ⊥ H ⊕ y Im H has three fixed directions ( y i , y j , y k ) (yi,yj,yk) ( y i , y j , y k ) . A connection on V V V splits the tangent sequence of its disk bundle, giving actual oriented identifications
T B = V R ⊕ R 3 , T N ≅ π ∗ ( 2 V R ⊕ R 3 ) . (169) TB=V_{\mathbb{R}}\oplus\mathbb{R}^3,\qquad TN\cong\pi^*(2V_{\mathbb{R}}\oplus\mathbb{R}^3). \tag*{(169)} T B = V R ⊕ R 3 , T N ≅ π ∗ ( 2 V R ⊕ R 3 ) . (169)
With Q = π ∗ P Q=\pi^*P Q = π ∗ P and its associated real quaternionic line Q H Q_{\mathbb{H}} Q H , these yield
E L = 2 Q H ⊕ R 8 ≅ T N ⊕ R 5 , E L ′ = R 16 . (170) E_L=2Q_{\mathbb{H}}\oplus\mathbb{R}^8\cong TN\oplus\mathbb{R}^5,\qquad E'_L=\mathbb{R}^{16}. \tag*{(170)} E L = 2 Q H ⊕ R 8 ≅ T N ⊕ R 5 , E L ′ = R 16 . (170)
The representation u ↦ diag ( L u , L u , I 8 ) u\mapsto\operatorname{diag}(L_u,L_u,I_8) u ↦ diag ( L u , L u , I 8 ) , where L u L_u L u is left quaternionic multiplication, lifts to Spin ( 16 ) \operatorname{Spin}(16) Spin ( 16 ) uniquely after fixing the identity because K K K is simply connected. These are therefore actual Spin gauge bundles. Since N N N retracts onto S 7 S^7 S 7 , both H 1 ( N ; Z / 2 ) H^1(N;\mathbb{Z}/2) H 1 ( N ; Z /2 ) and H 2 ( N ; Z / 2 ) H^2(N;\mathbb{Z}/2) H 2 ( N ; Z /2 ) vanish: there is no obstruction or choice of Spin structure in these identifications.
The unit-vector identification [ p , u ] ↦ p diag ( u , 1 ) [p,u]\mapsto p\operatorname{diag}(u,1) [ p , u ] ↦ p diag ( u , 1 ) identifies M M M with P P P . On this boundary, Q Q Q has the tautological section τ ( p ) = ( p , p ) \tau(p)=(p,p) τ ( p ) = ( p , p ) . To compare its frame with the left Lie frame, choose the principal connection α = pr k ( p − 1 d p ) \alpha=\operatorname{pr}_{\mathfrak{k}}(p^{-1}dp) α = pr k ( p − 1 d p ) , where the projection is onto the stabilizer Lie algebra, and write
X = ( α ζ − ζ ‾ γ ) ∈ s p ( 2 ) , α , γ ∈ Im H , ζ ∈ H . X=\begin{pmatrix}\alpha&\zeta\\-\overline{\zeta}&\gamma\end{pmatrix}\in\mathfrak{sp}(2),\qquad\alpha,\gamma\in\operatorname{Im}\mathbb{H},\quad\zeta\in\mathbb{H}. X = ( α − ζ ζ γ ) ∈ sp ( 2 ) , α , γ ∈ Im H , ζ ∈ H .
The vector p X pX pX projects to p ( ζ , γ ) p(\zeta,\gamma) p ( ζ , γ ) at p e 2 pe_2 p e 2 . Its horizontal V V V component is p e 1 ζ pe_1\zeta p e 1 ζ , and its vertical covariant derivative is p e 1 α pe_1\alpha p e 1 α . The outward normal supplies the remaining real direction p e 1 pe_1 p e 1 . Two copies of the τ \tau τ -frame, together with the three fixed directions, consequently give the stable left Lie frame up to a constant oriented basis change. Passing the outward normal across the seven base directions supplies the boundary minus sign; the resulting boundary order is ( α , ζ , γ ) (\alpha, \zeta, \gamma) ( α , ζ , γ ) . Fix one Spin lift of this frame on the simply connected group M M M and retain it in every boundary gluing. The diagonal identification in (C.2) then gives an integral nullhomotopy of λ ( T N ) − λ ( E L ) \lambda(TN)-\lambda(E_L) λ ( T N ) − λ ( E L ) whose boundary String structure is precisely s L s_L s L . At the differential level the boundary identification also uses the tautological gauge morphism γ ˇ τ \check{\gamma}_{\tau} γ ˇ τ of (A.5). Its three-form is zero in the framed flat model, but subtracting 2 γ ˇ τ 2\check{\gamma}_{\tau} 2 γ ˇ τ from the restricted disk trivialization retains the two specified gauge identifications.
The difference from the right frame can be measured integrally. Let U T h ∈ H 4 ( N , M ; Z ) U_{\mathrm{Th}} \in H^4(N,M;\mathbb{Z}) U Th ∈ H 4 ( N , M ; Z ) be the Thom class with disk-fibre integral + 1 +1 + 1 . Because H 3 ( N ; Z ) = H 4 ( N ; Z ) = 0 H^3(N;\mathbb{Z})=H^4(N;\mathbb{Z})=0 H 3 ( N ; Z ) = H 4 ( N ; Z ) = 0 , the connecting map is an isomorphism. Define ν ∈ H 3 ( M ; Z ) \nu\in H^3(M;\mathbb{Z}) ν ∈ H 3 ( M ; Z ) by δ ν = U T h \delta\nu=U_{\mathrm{Th}} δ ν = U Th ; its stabilizer-S 3 S^3 S 3 period is + 1 +1 + 1 . The relative classes in the specified frames are
c 2 ( Q , τ ) = − U T h , λ ( T N , f L ) = 2 U T h , λ ( T N , f R ) = − U T h , s R − s L = 3 ν . (171) \begin{aligned}
c_2(Q,\tau)&=-U_{\mathrm{Th}}, \qquad\lambda(TN,f_L)=2U_{\mathrm{Th}},\\
\lambda(TN,f_R)&=-U_{\mathrm{Th}}, \qquad s_R-s_L=3\nu. \tag*{(171)}
\end{aligned} c 2 ( Q , τ ) λ ( T N , f R ) = − U Th , λ ( T N , f L ) = 2 U Th , = − U Th , s R − s L = 3 ν . (171)
Here f L f_L f L , f R f_R f R are the left and right stable boundary frames, including the outward-normal convention. For the first equality, the tautological unit vector extends on a disk fibre by x ↦ x x\mapsto x x ↦ x , with one positive zero. Its relative Euler class is U T h U_{\mathrm{Th}} U Th , and the opposite complex orientation gives the minus sign in c 2 c_2 c 2 : the complex basis ( 1 , j ) (1,j) ( 1 , j ) for right multiplication by i i i induces the real basis ( 1 , i , j , − k ) (1,i,j,-k) ( 1 , i , j , − k ) . Each left multiplication block therefore contributes + U T h +U_{\mathrm{Th}} + U Th to λ \lambda λ ; the left stable transition has two such blocks.
On the stabilizer, a right-invariant vector X p X_p X p at p = diag ( u , 1 ) p=\operatorname{diag}(u,1) p = diag ( u , 1 ) has base projection ( ζ , γ ) (\zeta,\gamma) ( ζ , γ ) independent of u u u and vertical component α u \alpha u α u . Including the normal, the transition is R u ⊕ I 7 R_u\oplus I_7 R u ⊕ I 7 , with R u R_u R u denoting right multiplication. Quaternionic conjugation C H C_H C H satisfies R u = C H L u − 1 C H R_u=C_HL_{u^{-1}}C_H R u = C H L u − 1 C H . Constant conjugation preserves the trace polynomial, whereas inversion on S 3 S^3 S 3 has degree − 1 -1 − 1 . This gives the relative right-framed class − U T h -U_{\mathrm{Th}} − U Th . The unique ordinary String structure on N N N restricts to s b n d = s L + 2 ν = s R − ν s_{\mathrm{bnd}}=s_L+2\nu=s_R-\nu s bnd = s L + 2 ν = s R − ν , proving the last equality in (C.3). It compares the structures over the identity map; inversion below is a different comparison.
Smooth left connections and the selected local integral
The left construction needs a smooth bulk gauge connection that is flat in the tautological frame on a whole collar. In a local base gauge, write the fibre coordinate as x = r u x=ru x = r u and choose f ( r ) = r 2 g ( r 2 ) f(r)=r^2g(r^2) f ( r ) = r 2 g ( r 2 ) near zero, with g g g smooth and f = 1 f=1 f = 1 on the collar. Set
A = ( 1 − f ) a − f d u u − 1 . (172) A=(1-f)a-f\,du\,u^{-1}. \tag*{(172)} A = ( 1 − f ) a − f d u u − 1 . (172)
The identity r 2 d u u − 1 = Im ( d x x ˉ ) r^2du\,u^{-1}=\operatorname{Im}(dx\,\bar{x}) r 2 d u u − 1 = Im ( d x x ˉ ) proves smoothness at the centre. Under a base change p ↦ p k p\mapsto pk p ↦ p k , the coordinate becomes k − 1 u k^{-1}u k − 1 u , and A A A obeys the ordinary connection transformation law. In the τ \tau τ -frame it is ( 1 − f ) ( u − 1 a u + u − 1 d u ) (1-f)(u^{-1}au+u^{-1}du) ( 1 − f ) ( u − 1 a u + u − 1 d u ) , hence zero throughout the collar. The fixed Spin lift makes this the trivial Spin connection too; flatness alone would not specify that lift or the boundary gauge frame.
Choose a radial metric d r 2 + ℓ ( r ) 2 g S 3 dr^2+\ell(r)^2g_{S^3} d r 2 + ℓ ( r ) 2 g S 3 , with ℓ = r \ell=r ℓ = r near zero and ℓ \ell ℓ constant near the boundary. On the round base take A B ∗ = a V ⊕ d R 3 A_B^*=a_V\oplus d_{\mathbb{R}^3} A B ∗ = a V ⊕ d R 3 , for which λ ( A B ∗ ) = − c \lambda(A_B^*)=-c λ ( A B ∗ ) = − c , where c = c 2 ( P , a ) c=c_2(P,a) c = c 2 ( P , a ) denotes the characteristic form and a V a_V a V is induced by a a a . The vertical connection is fibre Levi–Civita vertically and uses horizontal K K K -transport otherwise. Its sum with π ∗ A B ∗ \pi^*A_B^* π ∗ A B ∗ is the product tangent connection A ∗ A_* A ∗ , preserving the outward normal on the collar. A positive radial isometry to (C.1) preserves the frame homotopy and identifies the intermediate connection as A 0 = π ∗ ( A B ∗ ⊕ a V ) A_0=\pi^*(A_B^*\oplus a_V) A 0 = π ∗ ( A B ∗ ⊕ a V ) , with the trivial stabilizing summands understood. Thus λ ( A 0 ) = − 2 π ∗ c \lambda(A_0)=-2\pi^*c λ ( A 0 ) = − 2 π ∗ c and λ ( E L , A ) = − 2 C \lambda(E_L,A)=-2C λ ( E L , A ) = − 2 C . This intermediate connection need not preserve the normal and is not used for gluing.
Transport on the same identified stable Spin bundle first from the gauge connection to A 0 A_0 A 0 and then to A ∗ A_* A ∗ . With C = c 2 ( Q , A ) C=c_2(Q,A) C = c 2 ( Q , A ) , the curving of the diagonal trivialization is
H L = 2 T q + K 3 , T q = T c 2 ( π ∗ a , A ) , K 3 = T λ ( A 0 , A ∗ ) , H_L=2T_q+K_3,\qquad T_q=T_{c_2}(\pi^*a,A),\qquad K_3=T_\lambda(A_0,A_*), H L = 2 T q + K 3 , T q = T c 2 ( π ∗ a , A ) , K 3 = T λ ( A 0 , A ∗ ) ,
d H L = 2 ( C − π ∗ c ) + λ ( A ∗ ) + 2 π ∗ c = λ ( A ∗ ) + 2 C . (173) dH_L=2(C-\pi^*c)+\lambda(A_*)+2\pi^*c=\lambda(A_*)+2C. \tag*{(173)} d H L = 2 ( C − π ∗ c ) + λ ( A ∗ ) + 2 π ∗ c = λ ( A ∗ ) + 2 C . (173)
For an invariant polynomial P \mathcal{P} P , the transgression convention is d T P ( A 0 , A 1 ) = P ( A 1 ) − P ( A 0 ) dT_{\mathcal{P}}(A_0,A_1)=\mathcal{P}(A_1)-\mathcal{P}(A_0) d T P ( A 0 , A 1 ) = P ( A 1 ) − P ( A 0 ) . This path retains the integral nullhomotopy, not only the displayed curvature equation. In the cochain notation of (10) , choose its integral trivializing cochain b b b . Then I ( H L ) − b − h X I(H_L)-b-h_X I ( H L ) − b − h X is a closed real three-cochain. Since H 3 ( N ; R ) = 0 H^3(N;\mathbb{R})=0 H 3 ( N ; R ) = 0 , it is δ k \delta k δ k for a real two-cochain k k k . The path and collar identifications retain the specified boundary morphisms, so all three trivialization equations hold. Redden’s torsor description explains why the integral nullhomotopy and its geometric refinement must both be retained [15 Propositions 3.11–3.12, Theorem 3.15] .
At A B ∗ A_B^* A B ∗ , the degree-four differential character λ ˇ + c ˇ \check{\lambda}+\check{c} λ ˇ + c ˇ is flat. Its flat class vanishes because H 3 ( S 7 ; R / Z ) = 0 H^3(S^7;\mathbb{R}/\mathbb{Z})=0 H 3 ( S 7 ; R / Z ) = 0 , and the integral base trivialization is unique. Choose its geometric representative with base three-form curving h = 0 h=0 h = 0 , keeping the characteristic morphisms and cochains. The matched base connections at both ends of the transgression give exact fibre-integrated form identities [7 §3, §6] :
π N ∗ C 2 = − c , π N ∗ ( T q C 2 ) = π N ∗ ( K 3 C 2 ) = 0 , a N , ∗ = ∫ N H L Y 8 = 0. (174) \pi_{N*}C^2=-c,\qquad\pi_{N*}(T_qC^2)=\pi_{N*}(K_3C^2)=0,\qquad a_{N,*}=\int_N H_LY_8=0. \tag*{(174)} π N ∗ C 2 = − c , π N ∗ ( T q C 2 ) = π N ∗ ( K 3 C 2 ) = 0 , a N , ∗ = ∫ N H L Y 8 = 0. (174)
Here π N ∗ \pi_{N*} π N ∗ integrates over the oriented disk fibres, and Y 8 = − C 2 / 3 Y_8=-C^2/3 Y 8 = − C 2 /3 includes the full contribution p 2 ( E L ) = 6 C 2 p_2(E_L)=6C^2 p 2 ( E L ) = 6 C 2 . In the universal K K K -equivariant disk, the Thom relation U T h = − c U T h U_{\mathrm{Th}}=-cU_{\mathrm{Th}} U Th = − c U Th and the relative equivariant Chern class C K = − U T h C_K=-U_{\mathrm{Th}} C K = − U Th give ∫ D 4 C K 2 = − c \int_{D^4}C_K^2=-c ∫ D 4 C K 2 = − c . The fibre integral lies in the Cartan complex of a point: the K K K -invariant polynomials in a formal Lie-algebra variable of degree two, with zero differential. The polynomial identity therefore becomes an equality of forms after association with a a a . The integrated products of the degree-three equivariant transgressions with C K 2 C_K^2 C K 2 have degree seven. The odd-degree part of that complex vanishes, proving the two zero pushforwards; the framed curvature is zero on the collar, so no boundary term is omitted. The vanishing concerns this chosen representative; the general local term remains ∫ B h C / 3 \int_B h_C/3 ∫ B h C /3 . For a finite change from the Levi–Civita base connection A B L C A_B^{LC} A B L C to A B ∗ A_B^* A B ∗ , set α = T λ ( A B L C , A B ∗ ) \alpha=T_\lambda(A_B^{LC},A_B^*) α = T λ ( A B L C , A B ∗ ) and transport the same trivialization by h α = h L C + α h_\alpha=h_{LC}+\alpha h α = h L C + α , with its morphism. For the associated mixed double of Section 4, Z = N ‾ ∪ M W r e f Z=\overline{N}\cup_M W_{\mathrm{ref}} Z = N ∪ M W ref reverses the disk and has trivial gauge curvature on the reference half. Writing π Z : Z → B \pi_Z:Z\to B π Z : Z → B for its projection, one therefore has π Z ∗ C 2 = c \pi_{Z*}C^2=c π Z ∗ C 2 = c . The index transgression is − π Z ∗ α C 2 / 3 -\pi_Z^*\alpha_C^2/3 − π Z ∗ α C 2 /3 , so
a N , ∗ − a N , L C = 1 3 ∫ B α C , μ ∗ = μ L C = − 1 3 ∫ B α C ( m o d Z ) . (175) a_{N,*}-a_{N,\mathrm{LC}}=\frac{1}{3}\int_B\alpha_C,\qquad\mu_*=\mu_{\mathrm{LC}}=-\frac{1}{3}\int_B\alpha_C\pmod{\mathbb{Z}}. \tag*{(175)} a N , ∗ − a N , LC = 3 1 ∫ B α C , μ ∗ = μ LC = − 3 1 ∫ B α C ( mod Z ) . (175)
Here both closed real eta values include the finite tangent correction of (B.17); they are not spectra of an additional torsionful operator. The changes cancel in ϕ = − a N − μ \phi=-a_N-\mu ϕ = − a N − μ . Thus the zero local term is a useful representative of the left datum evaluated in (84) , rather than an assumption that local terms always vanish.
A right tangent connection with positive unit period
For the right filling, allow a full Spin ( 16 ) \operatorname{Spin}(16) Spin ( 16 ) gauge connection on E R = T N ⊕ R 5 E_R=TN\oplus\mathbb{R}^5 E R = T N ⊕ R 5 , with E R ′ = R 16 E'_R=\mathbb{R}^{16} E R ′ = R 16 . Requiring the two-quaternionic-block reduction used on the left would impose an extra condition on this construction. The tangent connection can instead be controlled separately, so that its Pontryagin form vanishes pointwise on each disk fibre.
Choose metric connections D B , D V D^B,D^V D B , D V on T B , V R TB,V_R T B , V R and the horizontal distribution of D V D^V D V . Away from the zero section put n = x / ∣ x ∣ n=x/|x| n = x /∣ x ∣ and
K n = n ( D V n ) ∗ − ( D V n ) n ∗ , D ^ V = D V + κ K n . K_n=n(D^Vn)^*-(D^Vn)n^*,\qquad\widehat{D}^V=D^V+\kappa K_n. K n = n ( D V n ) ∗ − ( D V n ) n ∗ , D V = D V + κ K n .
where the star denotes the metric adjoint and the radial cutoff κ \kappa κ is zero near the centre and one on the boundary collar. The cutoff ensures smoothness, while K n n = − D V n K_nn=-D^Vn K n n = − D V n makes the normal parallel on the collar. For the radial metric with ℓ = r \ell=r ℓ = r near zero and ℓ = 1 \ell=1 ℓ = 1 on the collar, the positive oriented isometry J J J to π ∗ ( T B ⊕ V R ) \pi^*(TB\oplus V_R) π ∗ ( T B ⊕ V R ) fixes horizontal vectors, sends ∂ r \partial_r ∂ r to n n n and an angular vector r v rv r v to ℓ ( r ) v \ell(r)v ℓ ( r ) v . It is smooth at the centre and independent of the radial coordinate on the collar. Consequently
A T = J − 1 ( π ∗ D B ⊕ D ^ V ) J A_T=J^{-1}\left(\pi^*D^B\oplus\widehat{D}^V\right)J A T = J − 1 ( π ∗ D B ⊕ D V ) J
is a smooth metric Spin \operatorname{Spin} Spin connection with product, normal-preserving boundary behaviour. On a fixed fibre D V = d D^V=d D V = d , and the curvature entries in a fixed orthonormal frame are
F i j = d κ ∧ ( n i d n j − d n i n j ) + ( 2 κ − κ 2 ) d n i ∧ d n j . F_{ij}=d\kappa\wedge(n_i\,dn_j-dn_i\,n_j)+(2\kappa-\kappa^2)\,dn_i\wedge dn_j. F ij = d κ ∧ ( n i d n j − d n i n j ) + ( 2 κ − κ 2 ) d n i ∧ d n j .
Rotating n n n to e 0 e_0 e 0 at the point of calculation makes d n 0 = 0 dn_0=0 d n 0 = 0 and gives the nonzero entries as
F 0 a = κ ′ d r ∧ d n a , F a b = ( 2 κ − κ 2 ) d n a ∧ d n b , a , b ∈ { 1 , 2 , 3 } . F_{0a}=\kappa' dr\wedge dn_a,\qquad F_{ab}=(2\kappa-\kappa^2)dn_a\wedge dn_b,\qquad a,b\in\{1,2,3\}. F 0 a = κ ′ d r ∧ d n a , F ab = ( 2 κ − κ 2 ) d n a ∧ d n b , a , b ∈ { 1 , 2 , 3 } .
Each entry is decomposable, so tr R ( F 2 ) = − 2 ∑ i < j F i j ∧ F i j = 0 \operatorname{tr}_{\mathbb{R}}(F^2)=-2\sum_{i<j}F_{ij}\wedge F_{ij}=0 tr R ( F 2 ) = − 2 ∑ i < j F ij ∧ F ij = 0 . The base connection has no fibre curvature, and conjugation by J J J preserves the trace. Thus p 1 ( A T ) p_1(A_T) p 1 ( A T ) vanishes on every fibre even though the connection need not be fibrewise flat.
Positively orthonormalize the right Lie frame without changing its oriented Spin / String \operatorname{Spin}/\operatorname{String} Spin / String homotopy class. Extend its flat boundary gauge connection to A G A_G A G on the actual bundle E R E_R E R , constant on a collar; the affine space of connections permits such an extension. The diagonal Spin \operatorname{Spin} Spin identification fixes the integral refinement of
H R = T λ ( A G , A T ⊕ d R 5 ) , d H R = λ ( A T ) − λ ( E R , A G ) . (176) H_R=T_\lambda(A_G,A_T\oplus d_{\mathbb{R}^5}),\qquad dH_R=\lambda(A_T)-\lambda(E_R,A_G). \tag*{(176)} H R = T λ ( A G , A T ⊕ d R 5 ) , d H R = λ ( A T ) − λ ( E R , A G ) . (176)
The integral nullhomotopy and the real cochain completing this transgression are obtained as in (C.5), with their endpoint morphisms. Its boundary is s R s_R s R . The right relative class in (C.3) has disk integral − 1 -1 − 1 . Since the gauge connection is flat in that prescribed boundary frame, Stokes gives the real equality
∫ S 3 H R = ∫ D 4 ( λ ( A T ) − λ ( A G ) ) = 0 − ( − 1 ) = 1. (177) \int_{S^3} H_R = \int_{D^4} \left(\lambda(A_T)-\lambda(A_G)\right)=0-(-1)=1. \tag*{(177)} ∫ S 3 H R = ∫ D 4 ( λ ( A T ) − λ ( A G ) ) = 0 − ( − 1 ) = 1. (177)
The relative class is independent of a connection change with its fixed boundary trivialization; a boundary Chern–Simons term accounts for a change of representative. This argument uses the full Spin polynomial of (7) . The positive unit period is special to this right construction; the outward fibre period of the general induced sphere-bundle datum is − 2 -2 − 2 .
Inversion and the differential String cylinder
Inversion ι ( p ) = p − 1 \iota(p)=p^{-1} ι ( p ) = p − 1 exchanges the Lie frames through
d ι p ( p X ) = − X p − 1 . \mathrm{d}\iota_p(pX)=-Xp^{-1}. d ι p ( pX ) = − X p − 1 .
The constant matrix − I 10 -I_{10} − I 10 has positive determinant and is connected to I 10 I_{10} I 10 in S O ( 10 ) \mathrm{SO}(10) SO ( 10 ) . Keeping the outward normal fixed and fixing a Spin lift at one point therefore gives an oriented String identification
ι ∗ s R = s L . (178) \iota^*s_R=s_L. \tag*{(178)} ι ∗ s R = s L . (178)
This pullback relation coexists with the nonzero difference s R − s L = 3 ν s_R-s_L=3\nu s R − s L = 3 ν over the identity. Pull back the complete right metric, connections and refinement. Interpolate tangent metrics through positive bundle identifications, transporting the refinement by the corresponding Chern–Simons transgression. At a common tangent connection the integral lifts agree. Their remaining difference is a degree-three differential character with zero characteristic class, hence is represented by a global two-form β \beta β through (A.2). This fixes the remaining geometric difference without changing the integral lift. On a product-ended cylinder with interval coordinate s s s , choose χ \chi χ constant at zero and one near its respective ends. If H 0 H_0 H 0 is the initial endpoint curving pulled back to the product, the required three-form is
H c y l = H 0 + d ( χ β ) = H 0 + χ d β + χ d s ∧ β . (179) H_{\mathrm{cyl}}=H_0+\mathrm{d}(\chi\beta)=H_0+\chi\,\mathrm{d}\beta+\chi\,\mathrm{d}s\wedge\beta. \tag*{(179)} H cyl = H 0 + d ( χ β ) = H 0 + χ d β + χ d s ∧ β . (179)
The integral lift and endpoint cochain isomorphisms are retained. The interval term makes the added form exact on the cylinder and preserves the differential cocycle equation; interpolating only the endpoint three-forms would omit that term.
Both gauge connections on these cylinders are trivial in their specified frames. Their real virtual coefficient cancels as 256 − 128 − 128 = 0 256-128-128=0 256 − 128 − 128 = 0 , and Y 8 = 0 Y_8=0 Y 8 = 0 . The reference-and-boundary transport property of § 2 , with the real line conventions of Appendix § 14 , therefore identifies the two values [7 §2] . This complete transport, including the Spin lift and cochain morphisms, returns (84) to the right-framed positive-unit datum. Equality of bare periods would not supply it, and periods need not remain fixed during a change of tangent connection.
Finally, reversing orientation together with all associated boundary, differential and reference data reverses the local integral and the closed real eta contribution. Thus ϕ ( − M ) = − ϕ ( M ) \phi(-M)=-\phi(M) ϕ ( − M ) = − ϕ ( M ) and the response is complex conjugated, giving e − 2 π i / 3 e^{-2\pi i/3} e − 2 π i /3 for the reversed calibrated datum. Resetting its period independently to + 1 +1 + 1 would change the input. With the data retained as above, § 6 obtains the same e 2 π i / 3 e^{2\pi i/3} e 2 π i /3 response on the two specified Lie-framed data.
Associated families and the representation-valued index
The eleven-dimensional closed eta term can also be evaluated by reducing along the four-sphere fibres of the mixed double. This gives an analytic route to the intrinsic seven-dimensional expression in § 5 , separate from the twelve-dimensional Atiyah–Patodi–Singer (APS) evaluation in § 4 . The reduction retains the representation carried by the vertical zero modes and the connection induced on their bundle. It also retains a finite tangent transgression when the prescribed connection differs from the natural split connection. These two pieces are needed to recover the same differential-index value.
The associated-family reduction of Li and Sun [7 Appendices A–B] computes this term from an equivariant four-sphere index and its induced connection. The vertical zero modes specify the coefficient of an auxiliary base Dirac operator; they are not additional dynamical degrees of freedom. The vanishing eta form and finite connection transgression then identify the invariant computed by that operator. The conditional family in the final subsection instead concerns the three-sphere boundary of the disk fibres, whose spectral-section problem is separate from this closed calculation.
The real coefficient becomes an integer eta combination
Use the closed mixed double Z = N ‾ ∪ M W r e f = P × K F Z=\overline{N}\cup_M W_{\mathrm{ref}}=P\times_K F Z = N ∪ M W ref = P × K F , with K = Sp ( 1 ) K=\operatorname{Sp}(1) K = Sp ( 1 ) and F = S ( H ⊕ R ) ≅ H P 1 F=S(\mathbb{H}\oplus\mathbb{R})\cong\mathbb{H}P^1 F = S ( H ⊕ R ) ≅ H P 1 , constructed in § 4 . Its Spin structure and its coefficient line are the actual glued ones. The line on the lower disk has the left-multiplication action, whereas its frame on the upper disk has trivial action; the tautological boundary morphism glues them. In projective coordinates these are the restrictions of the tautological line on H P 1 \mathbb{H}P^1 H P 1 . The orientation is fixed by ∫ F c 2 ( U F ) = 1 \int_F c_2(U_F)=1 ∫ F c 2 ( U F ) = 1 , where U F U_F U F denotes its complex rank-two bundle. The invariant radial connection is fixed throughout the reduction. Thus the comparison uses the bundle, action, Spin lift and connection, rather than only a fibre characteristic number.
Restrict the actual real coefficient of § 2 to this K K K action. Let U Z U_Z U Z be the complex rank-two bundle underlying the glued quaternionic line on Z Z Z . In the representation identities, U U U denotes the defining complex rank-two representation of K K K and Ad R \operatorname{Ad}_{\mathbb{R}} Ad R its real adjoint representation. The restricted representations are
Δ R ∣ K = 16 H − 8 Ad R − 40 R , Δ C ∣ K = 32 U Z − 8 Ad ( U Z ) − 40. (180) \Delta_{\mathbb{R}|K}=16\mathbb{H}-8\operatorname{Ad}_{\mathbb{R}}-40\mathbb{R}, \qquad\Delta_{\mathbb{C}|K}=32U_Z-8\operatorname{Ad}(U_Z)-40. \tag*{(180)} Δ R ∣ K = 16 H − 8 Ad R − 40 R , Δ C ∣ K = 32 U Z − 8 Ad ( U Z ) − 40. (180)
One can see the multiplicities directly. The first vector representation complexifies to 4 U + 8 4U+8 4 U + 8 , and its tensor product with the trivial second vector is 64 U + 128 64U+128 64 U + 128 . For torus parameters ( x , x , x , x , 0 , 0 , 0 , 0 ) (x,x,x,x,0,0,0,0) ( x , x , x , x , 0 , 0 , 0 , 0 ) the first half-spin character is
128 cosh ( x / 2 ) 4 = 8 ( e 2 x + 1 + e − 2 x ) + 32 ( e x + e − x ) + 40. (181) 128\cosh(x/2)^4=8(e^{2x}+1+e^{-2x})+32(e^x+e^{-x})+40. \tag*{(181)} 128 cosh ( x /2 ) 4 = 8 ( e 2 x + 1 + e − 2 x ) + 32 ( e x + e − x ) + 40. (181)
It is therefore 8 Ad ( U ) + 32 U + 40 8\operatorname{Ad}(U)+32U+40 8 Ad ( U ) + 32 U + 40 ; the other half-spin character is 128 trivial copies. Subtraction gives the complex identity in (D.1). Since H C = U ⊕ U \mathbb{H}_{\mathbb C}=U\oplus U H C = U ⊕ U and the adjoint and trivial representations are real, this also gives the displayed actual virtual real identity. Unitary equivariant intertwiners induce parallel bundle identifications for the associated connections, including their specified boundary gluing.
The real normalization μ I = − ξ ( D Δ C ) / 2 \mu_I=-\xi(D_{\Delta_{\mathbb C}})/2 μ I = − ξ ( D Δ C ) /2 can consequently be rewritten as
μ L C ( Z ) = − 16 ξ ( D Z , U Z ) + 4 ξ ( D Z , Ad ( U Z ) ) + 20 ξ ( D Z , 1 ) ( m o d Z ) . (182) \mu_{\mathrm{LC}}(Z)=-16\xi(D_Z,U_Z)+4\xi(D_Z,\operatorname{Ad}(U_Z))+20\xi(D_Z,1)\pmod{\mathbb{Z}}. \tag*{(182)} μ LC ( Z ) = − 16 ξ ( D Z , U Z ) + 4 ξ ( D Z , Ad ( U Z )) + 20 ξ ( D Z , 1 ) ( mod Z ) . (182)
Here every ξ \xi ξ includes the kernel term, ξ ( D ) = ( η A P S ( D ) + dim C ker D ) / 2 \xi(D)=(\eta_{\mathrm{APS}}(D)+\dim_{\mathbb C}\ker D)/2 ξ ( D ) = ( η APS ( D ) + dim C ker D ) /2 . The real-valued representatives were combined before reduction modulo Z \mathbb{Z} Z . All subsequent quotient-valued family identities are used with the integer coefficients − 16 -16 − 16 , 4 4 4 , 20 20 20 ; none is divided by two. The stable tangent connection will be introduced through its transgression after the ordinary Levi–Civita operators have been evaluated.
The equivariant four-sphere index
The vertical index specifies which bundles twist the base Dirac operator. Its rank would specify only the net number of modes, losing their K K K action. To determine this action, first compute for the homogeneous G = Sp ( 2 ) G=\operatorname{Sp}(2) G = Sp ( 2 ) action on H P 1 \mathbb{H}P^1 H P 1 , and then restrict to the subgroup K = diag ( Sp ( 1 ) , 1 ) K=\operatorname{diag}(\operatorname{Sp}(1),1) K = diag ( Sp ( 1 ) , 1 ) . Write 4 4 4 for the defining complex representation of Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) , 1 1 1 for its trivial representation, and use undecorated integers for trivial multiplicities. The three complex indices are
Ind G D U F = − 1 , Ind G D Ad ( U F ) = − 4 , Ind G D 1 = 0. (183) \operatorname{Ind}_{G}D_{U_F}=-1,\qquad\operatorname{Ind}_{G}D_{\operatorname{Ad}(U_F)}=-4,\qquad\operatorname{Ind}_{G}D_1=0. \tag*{(183)} Ind G D U F = − 1 , Ind G D Ad ( U F ) = − 4 , Ind G D 1 = 0. (183)
The chirality and orientation fix the signs in this formula. Let Q F = U F ⊥ Q_F=U_F^\perp Q F = U F ⊥ in the equivariantly trivial quaternionic two-plane bundle. At [ 1 : 0 ] [1:0] [ 1 : 0 ] , the stabilizer Sp ( 1 ) a × Sp ( 1 ) b \operatorname{Sp}(1)_a\times\operatorname{Sp}(1)_b Sp ( 1 ) a × Sp ( 1 ) b acts on the tangent quaternion by q ↦ b q a − 1 q\mapsto bqa^{-1} q ↦ b q a − 1 . This is the Spin ( 4 ) \operatorname{Spin}(4) Spin ( 4 ) lift of the tangent action and supplies an honest equivariant Spin \operatorname{Spin} Spin bundle. The spin structure of S 4 S^4 S 4 is unique. Its two half-spin bundles are U F U_F U F and Q F Q_F Q F ; the identity
c 2 ( S − ) − c 2 ( S + ) = e ( T F ) , c 2 ( Q F ) = − c 2 ( U F ) , ∫ F e ( T F ) = 2 (184) c_2(S^-)-c_2(S^+)=e(TF),\qquad c_2(Q_F)=-c_2(U_F),\qquad\int_F e(TF)=2 \tag*{(184)} c 2 ( S − ) − c 2 ( S + ) = e ( T F ) , c 2 ( Q F ) = − c 2 ( U F ) , ∫ F e ( T F ) = 2 (184)
then identifies S − = U F S^-=U_F S − = U F and S + = Q F S^+=Q_F S + = Q F in the chosen orientation.
For a regular torus element t = diag ( a , b ) t=\operatorname{diag}(a,b) t = diag ( a , b ) , set A = a + a − 1 A=a+a^{-1} A = a + a − 1 and B = b + b − 1 B=b+b^{-1} B = b + b − 1 . If a ≠ b , b − 1 a\ne b,b^{-1} a = b , b − 1 , its only fixed quaternionic lines are p = [ 1 : 0 ] p=[1:0] p = [ 1 : 0 ] and q = [ 0 : 1 ] q=[0:1] q = [ 0 : 1 ] . A line with both coordinates nonzero would require a a a and b b b to be conjugate unit quaternions, excluded by this condition. The tangent weights give
det R ( 1 − t ∣ T F ) = ( 1 − a b ) ( 1 − ( a b ) − 1 ) ( 1 − a / b ) ( 1 − b / a ) = ( A − B ) 2 . (185) \det_{\mathbb{R}}(1-t|_{TF})=(1-ab)(1-(ab)^{-1})(1-a/b)(1-b/a)=(A-B)^2. \tag*{(185)} R det ( 1 − t ∣ T F ) = ( 1 − ab ) ( 1 − ( ab ) − 1 ) ( 1 − a / b ) ( 1 − b / a ) = ( A − B ) 2 . (185)
The coefficient and spin supercharacters at these points are
Table 1.
The elliptic Dirac complex is geometric under the compact-group action, and the fixed points are transverse by (D.6). Thus the Atiyah–Bott fixed-point formula applies [26 §2, Theorem A, (2.8)–(2.9)] . To match its pullback convention, use the base map t − 1 t^{-1} t − 1 and the fibre lift induced by t t t ; the real determinant in the denominator is unchanged. For coefficient characters f ( a ) f(a) f ( a ) , f ( b ) f(b) f ( b ) , the sum of its two contributions is
( B − A ) f ( a ) + ( A − B ) f ( b ) ( A − B ) 2 = f ( b ) − f ( a ) A − B . (186) \frac{(B-A)f(a)+(A-B)f(b)}{(A-B)^2}=\frac{f(b)-f(a)}{A-B}. \tag*{(186)} ( A − B ) 2 ( B − A ) f ( a ) + ( A − B ) f ( b ) = A − B f ( b ) − f ( a ) . (186)
The three choices f ( a ) = A f(a)=A f ( a ) = A , A 2 − 1 A^2-1 A 2 − 1 , 1 1 1 give − 1 -1 − 1 , − ( A + B ) -(A+B) − ( A + B ) , 0 0 0 . These are the characters in (D.4). Continuity extends equality from the regular torus to the whole torus, and conjugation extends it to G G G ; virtual characters determine the representation ring.
The ordinary twisted index provides a separate sign and rank check. Since p 1 ( T S 4 ) = 0 p_1(TS^4)=0 p 1 ( T S 4 ) = 0 , the degree-four terms ch ( U F ) = 2 − c 2 ( U F ) + ⋯ \operatorname{ch}(U_F)=2-c_2(U_F)+\cdots ch ( U F ) = 2 − c 2 ( U F ) + ⋯ and ch ( Ad ( U F ) ) = 3 − 4 c 2 ( U F ) + ⋯ \operatorname{ch}(\operatorname{Ad}(U_F))=3-4c_2(U_F)+\cdots ch ( Ad ( U F )) = 3 − 4 c 2 ( U F ) + ⋯ give ranks − 1 -1 − 1 , − 4 -4 − 4 , 0 0 0 [27 §5, Theorem (5.3), Remark 2] . They do not determine the representation. In particular, the central element − I -I − I acts by − 1 -1 − 1 on the spinors and by + 1 +1 + 1 on the adjoint twist. The virtual adjoint-twist index has character + 4 +4 + 4 there, as does − 4 -4 − 4 ; − 4 1 -4^1 − 4 1 would have character − 4 -4 − 4 . Replacing one by the other would therefore change the bundle entering the base eta invariant.
Combining (D.4) with the integer coefficients in (D.3) gives
R e f f = 16 − 44 , R e f f ∣ K = 8 − 4 U , 4 ∣ K = U + 2. (187) R_{\mathrm{eff}}=16-44,\qquad R_{\mathrm{eff}}\big|_K=8-4U,\qquad4\big|_K=U+2. \tag*{(187)} R eff = 16 − 44 , R eff K = 8 − 4 U , 4 K = U + 2. (187)
The homogeneous metric and connection can be deformed to the invariant radial data through K K K -invariant positive metrics and unitary connections. The equivariant elliptic index is unchanged. This identifies its virtual representation, but the connection on the actual graded kernels still has to be identified before using (D.8) analytically.
The rank-zero result is particularly instructive here. Both 16 − 44 16-44 16 − 44 and its restriction 8 − 4 U 8-4U 8 − 4 U have virtual rank zero, but the latter has degree-four Chern character 4 c 4c 4 c , where c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) . It retains the quaternionic background in the reduced calculation. A rank-only replacement would produce the zero coefficient and erase that dependence. Conversely, the representation-valued index is a difference of graded kernels, not a statement that the individual kernels vanish or that nonzero vertical eigenvalues may be omitted from an arbitrary real eta formula. The associated-family argument determines precisely the quotient-valued invariant to which this virtual coefficient applies.
The relevant connection statement holds for an associated family more generally. Let a compact connected Lie group K K K act isometrically on a closed spin manifold F F F of positive even dimension, with an honest Spin lift. Fix an equivariant Hermitian twisting bundle E F \mathcal{E}_F E F with invariant unitary connection. For a principal right bundle ( P , a ) (P,a) ( P , a ) on a closed odd-dimensional spin manifold B B B , let π : Z → B \pi:Z\to B π : Z → B be the associated bundle projection and use
Z = P × K F , E Z = P × K E F , ( p k , f ) ∼ ( p , k f ) , g ε = g F + ε − 2 π ∗ g B . (188) Z=P\times_K F,\qquad\mathcal{E}_Z=P\times_K\mathcal{E}_F,\qquad(pk,f)\sim(p,kf),\qquad g_\varepsilon=g_F+\varepsilon^{-2}\pi^*g_B. \tag*{(188)} Z = P × K F , E Z = P × K E F , ( p k , f ) ∼ ( p , k f ) , g ε = g F + ε − 2 π ∗ g B . (188)
The horizontal distribution is induced by a a a . Horizontal transport on the twist is the prescribed K K K action, and the tangent Spin structure is the product one for T Z = T V Z ⊕ π ∗ T B TZ = T^{V}Z \oplus\pi^{*}TB T Z = T V Z ⊕ π ∗ T B , where T V Z T^{V}Z T V Z is the tangent bundle along the fibres. At unit horizontal scale the Clifford convention is c Z ( v ) = c F ( v ) ⊗ 1 c_Z(v) = c_F(v) \otimes1 c Z ( v ) = c F ( v ) ⊗ 1 and c Z ( X H ) = Γ F ⊗ c B ( X ) c_Z(X^H) = \Gamma_F \otimes c_B(X) c Z ( X H ) = Γ F ⊗ c B ( X ) , where Γ F \Gamma_F Γ F is vertical chirality, with eigenspaces S F ± S_F^{\pm} S F ± , and X H X^H X H is the horizontal lift of a base tangent vector X X X .
Let R ( K ) R(K) R ( K ) denote the complex representation ring of K K K . For the fixed fibre operator define
V ± = ker ( D F ∣ S F ± ⊗ E F ) , V ± = P × K V ± , R = V + − V − ∈ R ( K ) . (189) V^{\pm} = \ker(D_F|_{S_F^{\pm}\otimes E_F}), \qquad\mathcal{V}^{\pm} = P \times_K V^{\pm}, \qquad R = V^{+} - V^{-} \in R(K). \tag*{(189)} V ± = ker ( D F ∣ S F ± ⊗ E F ) , V ± = P × K V ± , R = V + − V − ∈ R ( K ) . (189)
Horizontal transport is by spin isometries preserving the twisting connection and vertical volume. The fibres are totally geodesic, with zero mean curvature. The vertical connection is fibre Levi–Civita on vertical arguments and satisfies ∇ X H V U = [ X H , U ] \nabla^{V}_{X^H}U = [X^H,U] ∇ X H V U = [ X H , U ] on horizontal arguments, for a vertical vector field U U U . Write ρ \rho ρ for the K K K action on fibre spinors. The induced connection on their Hilbert bundle is consequently
∇ H = d + ρ ∗ ( a ) , [ ∇ H , D F ] = 0. (190) \nabla^{\mathcal{H}} = d + \rho_{*}(a), \qquad[\nabla^{\mathcal{H}},D_F] = 0. \tag*{(190)} ∇ H = d + ρ ∗ ( a ) , [ ∇ H , D F ] = 0. (190)
The second identity follows from equivariance of the fixed operator. Thus the kernel projections are parallel, and their projected connections are exactly the associated connections on V ± \mathcal{V}^{\pm} V ± . These are the projected kernel connections entering the analytic index [28, §3.1, (3.7)–(3.10), p. 919]. The vertical kernels form smooth bundles of constant rank. This is a geometric consequence of the specified associated transport, rather than a separate kernel-stability assumption for the total operator.
A section of V ± \mathcal{V}^{\pm} V ± is transported by the representation of K K K on the fixed kernel of D F D_F D F . Its covariant derivative is therefore induced by the original connection a a a , and the parallel kernel projection leaves this derivative unchanged. An equivariant intertwiner consequently identifies the twisted base operators with their connections. Equality of characteristic classes alone would not give this spectral identification.
The same geometry makes the Bismut–Cheeger eta form identically zero. Set T ( X , Y ) = − pr V [ X H , Y H ] \mathcal{T}(X,Y) = -\operatorname{pr}_{V}[X^H,Y^H] T ( X , Y ) = − pr V [ X H , Y H ] and C = c F ( T ) C = c_F(\mathcal{T}) C = c F ( T ) . For the right-bundle convention above, T = ( Ω a ) F \mathcal{T} = (\Omega_a)_F T = ( Ω a ) F , where Ω a \Omega_a Ω a is the curvature of a a a and the subscript F F F denotes its induced vertical vector field. Equivariance of b ↦ c F ( b F ) b \mapsto c_F(b_F) b ↦ c F ( b F ) sends the Bianchi identity to d ∇ E C = 0 d^{\nabla^{\mathcal{E}}}C = 0 d ∇ E C = 0 . This is an exterior covariant derivative identity, not directional covariant constancy of the curvature. With the geometric superconnection convention of [29 §1.1, pp. 271–273] ,
A t = t D F + ∇ H − C 4 t , (191) A_t = \sqrt{t}D_F + \nabla^{\mathcal{H}} - \frac{C}{4\sqrt{t}}, \tag*{(191)} A t = t D F + ∇ H − 4 t C , (191)
A t 2 = t D F 2 + ( ∇ H ) 2 − 1 4 ( D F C + C D F ) + C 2 16 t , t > 0. (192) A_t^2 = tD_F^2 + (\nabla^{\mathcal{H}})^2 - \frac{1}{4}(D_FC + CD_F) + \frac{C^2}{16t}, \qquad t > 0. \tag*{(192)} A t 2 = t D F 2 + ( ∇ H ) 2 − 4 1 ( D F C + C D F ) + 16 t C 2 , t > 0. (192)
Every coefficient of the square is even in vertical chirality and has even form degree. In contrast, ∂ t A t = D F / ( 2 t ) + C / ( 8 t 3 / 2 ) \partial_tA_t = D_F/(2\sqrt{t}) + C/(8t^{3/2}) ∂ t A t = D F / ( 2 t ) + C / ( 8 t 3/2 ) has chirality-odd coefficients. The heat operator is smoothing for t > 0 t > 0 t > 0 , so the actual fibrewise supertrace vanishes:
Str ( ( ∂ t A t ) e − A t 2 ) = 0 , η ~ = 0. (193) \operatorname{Str}\bigl((\partial_tA_t)e^{-A_t^2}\bigr) = 0, \qquad\widetilde{\eta} = 0. \tag*{(193)} Str ( ( ∂ t A t ) e − A t 2 ) = 0 , η = 0. (193)
The vanishing holds in every degree of the defining eta-form integral. It is stronger than a cohomological vanishing and uses the natural horizontal connection. Isotriviality of the fibres by itself would not give this conclusion for arbitrary horizontal data.
In particular, the proof did not assume that a a a is flat. Its curvature appears in C C C and in ( ∇ H ) 2 (\nabla^{\mathcal{H}})^2 ( ∇ H ) 2 , but the chirality of the heat-supertrace integrand still forces its vanishing. Nor did the proof use a vanishing odd cohomology group on the base. The eta form is zero as the actual differential form determined by this geometry, so its pairing in the adiabatic formula vanishes before any cohomological simplification. This distinction matters when the answer is subsequently compared at specified finite connections.
The reduced limit and the finite tangent correction
The smooth-kernel reduced adiabatic formula of Freed–Lott [28 Theorem 6.2, (6.15), p. 938] now gives
lim ε → 0 [ ξ ( D Z , g ε ) ] = [ ξ ( B , R ) ] + [ ∫ B A ^ ( T B ) η ~ ] = [ ξ ( B , R ) ] in R / Z , (194) \lim_{\varepsilon\to0}[\xi(D_{Z,g_\varepsilon})]=[\xi(B,R)]+\left[\int_B\widehat{A}(TB)\widetilde{\eta}\right]=[\xi(B,R)]\ \text{in }\mathbb{R}/\mathbb{Z}, \tag*{(194)} ε → 0 lim [ ξ ( D Z , g ε )] = [ ξ ( B , R )] + [ ∫ B A ( T B ) η ] = [ ξ ( B , R )] in R / Z , (194)
where ξ ( B , R ) = ξ ( D B , V + ) − ξ ( D B , V − ) \xi(B,R)=\xi(D_{B,V^+})-\xi(D_{B,V^-}) ξ ( B , R ) = ξ ( D B , V + ) − ξ ( D B , V − ) . The Spin case is the cited Spin c \operatorname{Spin}^c Spin c formula with trivial determinant connection. Its fibre-shrinking metric is ε 2 g ε = ε 2 g F + π ∗ g B \varepsilon^2g_\varepsilon=\varepsilon^2g_F+\pi^*g_B ε 2 g ε = ε 2 g F + π ∗ g B , after a positive parameter change. A uniform positive rescaling multiplies all nonzero eigenvalues by the same positive factor. The eta function acquires the corresponding factor to the power of its argument, which is one at zero, and the kernel dimension is unchanged. Thus the two metric conventions have the same reduced invariant. Compactness, the even-dimensional fibres, their Spin lift, and the smooth associated kernels supply the required hypotheses. Invertibility of the effective base operator and eventual constancy of the total-space kernel are not required for this quotient-valued limit. In particular, we do not use the stronger real-valued formula of Dai’s Theorem 0.1 with its additional total-kernel hypothesis.
This is a limiting identity, not a finite-ε \varepsilon ε error estimate. The effective operator is D B , V + ⊕ ( − D B , V − ) D_{B,V^+}\oplus(-D_{B,V^-}) D B , V + ⊕ ( − D B , V − ) ; its reduced invariant differs from ξ ( B , R ) \xi(B,R) ξ ( B , R ) by dim ker D B , V − \dim\ker D_{B,V^-} dim ker D B , V − , an integer. Small eigenvalues and zero modes cannot simply be dropped in a real-valued formula. For the flat product T 2 × S 1 T^2\times S^1 T 2 × S 1 with parallel spinors and trivial twist, the graded vertical index is zero while the total reduced eta is 1. This product is consistent with (D.15), but rules out an unqualified real equality. The limit supplies no half-integral refinement of a class in R / Z \mathbb{R}/\mathbb{Z} R / Z .
To return to a fixed metric, one must compare the tangent connections. Put ∇ n a t ⊕ = ∇ V ⊕ π ∗ ∇ B , L C \nabla^\oplus_{\mathrm{nat}}=\nabla^V\oplus\pi^*\nabla^{B,\mathrm{LC}} ∇ nat ⊕ = ∇ V ⊕ π ∗ ∇ B , LC , and retain the fixed twisting connection. With d CS P ( ∇ 0 , ∇ 1 ) = P ( ∇ 1 ) − P ( ∇ 0 ) d\,\operatorname{CS}_{P}(\nabla_0,\nabla_1)=P(\nabla_1)-P(\nabla_0) d CS P ( ∇ 0 , ∇ 1 ) = P ( ∇ 1 ) − P ( ∇ 0 ) , the finite formula is
[ ξ ( D Z , g ) ] = [ ξ ( B , R ) ] + [ ∫ Z CS A ^ ( ∇ n a t ⊕ , ∇ g Z , L C ) ch ( ∇ E Z ) ] . (195) [\xi(D_{Z,g})]=[\xi(B,R)]+\left[\int_Z\operatorname{CS}_{\widehat{A}}(\nabla^\oplus_{\mathrm{nat}},\nabla^{Z,\mathrm{LC}}_g)\operatorname{ch}(\nabla^{E_Z})\right]. \tag*{(195)} [ ξ ( D Z , g )] = [ ξ ( B , R )] + [ ∫ Z CS A ( ∇ nat ⊕ , ∇ g Z , LC ) ch ( ∇ E Z ) ] . (195)
Only the total-space dimension component is integrated. We recall why the limiting connection in this formula is the natural split one. In the fixed unrescaled splitting, Koszul’s formula gives a smooth limit
∇ g ε Z , L C ⟶ ∇ n a t ⊕ + N , (196) \nabla^{Z,\mathrm{LC}}_{g_\varepsilon}\longrightarrow\nabla^\oplus_{\mathrm{nat}}+\mathcal{N}, \tag*{(196)} ∇ g ε Z , LC ⟶ ∇ nat ⊕ + N , (196)
where N \mathcal{N} N has only the upper block π ∗ T B → T V Z \pi^*TB \to T^VZ π ∗ T B → T V Z ; the opposite block is O ( ε 2 ) O(\varepsilon^2) O ( ε 2 ) . The diagonal limits are the vertical and base connections already specified. Along ∇ n a t ⊕ + s N \nabla_{\mathrm{nat}}^{\oplus}+s\mathcal{N} ∇ nat ⊕ + s N , curvature is upper triangular with fixed diagonal. A trace with one inserted N \mathcal{N} N therefore vanishes. The Pontryagin trace polynomials defining each relevant degree of A ^ \widehat{A} A imply
CS A ^ ( ∇ n a t ⊕ , ∇ n a t ⊕ + N ) = 0 , CS A ^ ( ∇ n a t ⊕ , ∇ g ε Z , L C ) ⟶ 0. (197) \operatorname{CS}_{\widehat{A}}(\nabla_{\mathrm{nat}}^{\oplus},\nabla_{\mathrm{nat}}^{\oplus}+\mathcal{N})=0,\qquad\operatorname{CS}_{\widehat{A}}(\nabla_{\mathrm{nat}}^{\oplus},\nabla_{g_\varepsilon}^{Z,\mathrm{LC}})\longrightarrow0. \tag*{(197)} CS A ( ∇ nat ⊕ , ∇ nat ⊕ + N ) = 0 , CS A ( ∇ nat ⊕ , ∇ g ε Z , LC ) ⟶ 0. (197)
The second convergence is smooth because the finite-degree transgression is polynomial in connection coefficients and their derivatives. The limit need not be metric; the same invariant trace polynomials apply to that linear connection. The block estimate here identifies the connection limit, without asserting a spectral remainder estimate.
For a homogeneous invariant polynomial P P P of degree j j j , set ∇ s = ∇ 0 + s ( ∇ 1 − ∇ 0 ) \nabla_s=\nabla_0+s(\nabla_1-\nabla_0) ∇ s = ∇ 0 + s ( ∇ 1 − ∇ 0 ) and let F s F_s F s be its curvature. The transgression is j ∫ 0 1 P ( ∇ 1 − ∇ 0 , F s , … , F s ) d s j\int_0^1P(\nabla_1-\nabla_0,F_s,\ldots,F_s)\,ds j ∫ 0 1 P ( ∇ 1 − ∇ 0 , F s , … , F s ) d s [24 Proposition 3.8] . Its triangle identity follows by integrating the closed characteristic form of the affine connection on Z × Δ 2 Z\times\Delta^2 Z × Δ 2 and applying Stokes with boundary [ 1 , 2 ] − [ 0 , 2 ] + [ 0 , 1 ] [1,2]-[0,2]+[0,1] [ 1 , 2 ] − [ 0 , 2 ] + [ 0 , 1 ] . The oriented two-simplex integral S P S_P S P obeys
CS P ( ∇ 0 , ∇ 1 ) + CS P ( ∇ 1 , ∇ 2 ) − CS P ( ∇ 0 , ∇ 2 ) = d S P . (198) \operatorname{CS}_P(\nabla_0,\nabla_1)+\operatorname{CS}_P(\nabla_1,\nabla_2)-\operatorname{CS}_P(\nabla_0,\nabla_2)=dS_P. \tag*{(198)} CS P ( ∇ 0 , ∇ 1 ) + CS P ( ∇ 1 , ∇ 2 ) − CS P ( ∇ 0 , ∇ 2 ) = d S P . (198)
The two-simplex identity, rather than the infinitesimal proposition alone, justifies inserting the intermediate connection. Exact terms integrate to zero on closed Z Z Z against the closed twisting character.
Finally, twisted APS on a product-collar cylinder with boundary Z 1 ⊔ ( − Z 0 ) Z_1\sqcup(-Z_0) Z 1 ⊔ ( − Z 0 ) gives
[ ξ ( D Z , g 1 ) − ξ ( D Z , g 0 ) ] = ∫ Z CS A ^ ( ∇ g 0 Z , L C , ∇ g 1 Z , L C ) ch ( ∇ E Z ) (199) [\xi(D_{Z,g_1})-\xi(D_{Z,g_0})]=\int_Z\operatorname{CS}_{\widehat{A}}(\nabla_{g_0}^{Z,\mathrm{LC}},\nabla_{g_1}^{Z,\mathrm{LC}})\operatorname{ch}(\nabla^{E_Z}) \tag*{(199)} [ ξ ( D Z , g 1 ) − ξ ( D Z , g 0 )] = ∫ Z CS A ( ∇ g 0 Z , LC , ∇ g 1 Z , LC ) ch ( ∇ E Z ) (199)
[8 Theorem (4.2), (4.3)] . The endpoint identity ξ ( − D ) = − ξ ( D ) + dim ker D \xi(-D)=-\xi(D)+\dim\ker D ξ ( − D ) = − ξ ( D ) + dim ker D retains the integer kernel contribution even when an endpoint is not invertible. Set g 0 = g ε g_0=g_\varepsilon g 0 = g ε , insert ∇ n a t ⊕ \nabla_{\mathrm{nat}}^{\oplus} ∇ nat ⊕ using (D.19), and take the reduced limit. Equations (D.15) and (D.18) give (D.16), including its sign.
For a specified split stable metric connection ∇ ∗ \nabla_* ∇ ∗ on the same stabilized tangent bundle, define the differential-index value by
ξ ∗ ( Z , E Z ) : = [ ξ ( D Z , g ) ] − [ ∫ Z CS A ^ ( ∇ ∗ , ∇ g Z , L C ) ch ( ∇ E Z ) ] = [ ξ ( B , R ) ] + [ ∫ Z CS A ^ ( ∇ n a t ⊕ , ∇ ∗ ) ch ( ∇ E Z ) ] . (200) \begin{aligned}
\xi_*(Z,E_Z):=[\xi(D_{Z,g})]-\left[\int_Z\operatorname{CS}_{\widehat{A}}(\nabla_*,\nabla_g^{Z,\mathrm{LC}})\operatorname{ch}(\nabla^{E_Z})\right] \\
&=[\xi(B,R)]+\left[\int_Z\operatorname{CS}_{\widehat{A}}(\nabla_{\mathrm{nat}}^{\oplus},\nabla_*)\operatorname{ch}(\nabla^{E_Z})\right]. \tag*{(200)}
\end{aligned} ξ ∗ ( Z , E Z ) := [ ξ ( D Z , g )] − [ ∫ Z CS A ( ∇ ∗ , ∇ g Z , LC ) ch ( ∇ E Z ) ] = [ ξ ( B , R )] + [ ∫ Z CS A ( ∇ nat ⊕ , ∇ ∗ ) ch ( ∇ E Z ) ] . (200)
The triangle identity and cylinder variation make it independent of the auxiliary metric g g g . It is not the raw eta invariant of a torsionful tangent connection. A change of twisting connection would require its own Chern–Simons contribution; the twist was fixed here.
For example, if two auxiliary metrics are used in the first line of (D.21), their ordinary reduced eta values differ by the cylinder integral (D.20). The difference between their two subtracted transgressions is that same integral by (D.19), modulo an exact form. These differences cancel on the closed total space. The result still depends on the specified stable connection and twist; independence of the auxiliary Levi–Civita representative does not remove those background data from the differential-index value.
The base response and the special stable connection
For the four-sphere family, the kernel representation identified above has precisely the associated connection required by (D.21). Equality of virtual representations is realized, by complete reducibility, as an equivariant stable isomorphism after adding the same representation to both sides. A unitary intertwiner becomes parallel under association to ( P , a ) (P,a) ( P , a ) . Additivity of reduced eta, including the kernels, therefore permits the effective coefficient − 4 U + 8 -4U+8 − 4 U + 8 , rather than only its rank zero.
At the general natural split tangent connection, the transgression in (D.21) vanishes. Applying that formula to the three twists with the integer multiplicities gives
μ ∗ ( Z ) = ξ ( B , − 4 U + 8 ) in R / Z . (201) \mu_*(Z)=\xi(B,-4U+8)\quad\text{in }\mathbb{R}/\mathbb{Z}. \tag*{(201)} μ ∗ ( Z ) = ξ ( B , − 4 U + 8 ) in R / Z . (201)
Combining it with a N = ∫ B h ∧ c / 3 a_N=\int_B h\wedge c/3 a N = ∫ B h ∧ c /3 and the scalar sign ϕ = − a N − μ ∗ \phi=-a_N-\mu_* ϕ = − a N − μ ∗ yields
ϕ = ξ ( B , 4 U − 8 ) − 1 3 ∫ B h ∧ c = K 3 in R / Z . (202) \phi=\xi(B,4U-8)-\frac{1}{3}\int_B h\wedge c=\frac{K}{3}\quad\text{in }\mathbb{R}/\mathbb{Z}. \tag*{(202)} ϕ = ξ ( B , 4 U − 8 ) − 3 1 ∫ B h ∧ c = 3 K in R / Z . (202)
This is the intrinsic expression and relative pairing already established in § 5 . The equality needs no rational H 3 H^3 H 3 or H 4 H^4 H 4 vanishing; those restrictions enter only its Crowley–Goette comparison. The local term remains generally nonzero.
The integral differential trivialization on the base is still present in this formula, through its curving h h h and the relative class defining K K K . The family theorem evaluates the closed spectral part; it does not trivialize λ + c \lambda+c λ + c over an eight-dimensional spin/gauge filling or replace the specified integral nullhomotopy by a freely chosen real primitive. Thus the reduction and the relative pairing have complementary roles: the former produces the base operator and its connection, while the latter fixes the globally quantized combination with the disk term.
On the original S 7 S^7 S 7 base, the special splitting T B = V R ⊕ R 3 TB=V_{\mathbb{R}}\oplus\mathbb{R}^3 T B = V R ⊕ R 3 supplies A B ∗ = a V ⊕ d R 3 A_B^*=a_V\oplus d_{\mathbb{R}^3} A B ∗ = a V ⊕ d R 3 . Let k 3 = T λ ( A B L C , A B ∗ ) k_3=T_\lambda(A_B^{\mathrm{LC}},A_B^*) k 3 = T λ ( A B LC , A B ∗ ) and keep the same vertical connection on Z Z Z . The index polynomial governing the eleven-dimensional eta invariant is I 12 = − ( λ ( T Z ) + 2 C ) C 2 / 3 I_{12}=-(\lambda(TZ)+2C)C^2/3 I 12 = − ( λ ( T Z ) + 2 C ) C 2 /3 , with C = c 2 ( U Z ) C=c_2(U_Z) C = c 2 ( U Z ) . Its tangent dependence is linear in λ \lambda λ , so the natural-to-stable transgression is exactly
T I ( A n a t , A ∗ ) = − 1 3 π ∗ k 3 C 2 , π Z ∗ C 2 = c . (203) T_I(A_{\mathrm{nat}},A_*)=-\frac{1}{3}\pi^*k_3C^2,\qquad\pi_Z^*C^2=c. \tag*{(203)} T I ( A nat , A ∗ ) = − 3 1 π ∗ k 3 C 2 , π Z ∗ C 2 = c . (203)
The pushforward is the exact form identity for the glued radial data of § 4 and § 6 . It gives
μ ∗ ( Z ) = ξ e f f L C ( B ) − 1 3 ∫ B k 3 ∧ c = ξ e f f ∗ ( B ) , ξ e f f L C ( B ) = ξ ( B , − 4 U + 8 ) . (204) \mu_*(Z)=\xi_{\mathrm{eff}}^{\mathrm{LC}}(B)-\frac{1}{3}\int_B k_3\wedge c=\xi_{\mathrm{eff}}^*(B),\qquad\xi_{\mathrm{eff}}^{\mathrm{LC}}(B)=\xi(B,-4U+8). \tag*{(204)} μ ∗ ( Z ) = ξ eff LC ( B ) − 3 1 ∫ B k 3 ∧ c = ξ eff ∗ ( B ) , ξ eff LC ( B ) = ξ ( B , − 4 U + 8 ) . (204)
Here ξ e f f ∗ ( B ) \xi_{\mathrm{eff}}^*(B) ξ eff ∗ ( B ) denotes the base differential-index value at A B ∗ A_B^* A B ∗ , obtained from its ordinary reduced eta invariant by the specified tangent transgression.
The equality with the stable base value can be checked from its full eight-form, rather than inferred from the fibre index. Use roots x , − x , y , − y x,-x,y,-y x , − x , y , − y for 4 4 4 , set c 2 = − ( x 2 + y 2 ) c_2=-(x^2+y^2) c 2 = − ( x 2 + y 2 ) and T 4 = 2 ( x 4 + y 4 ) T_4=2(x^4+y^4) T 4 = 2 ( x 4 + y 4 ) , and expand ch ( 4 ) = 4 − c 2 + T 4 / 24 + ⋯ \operatorname{ch}(4)=4-c_2+T_4/24+\cdots ch ( 4 ) = 4 − c 2 + T 4 /24 + ⋯ . Since R e f f R_{\mathrm{eff}} R eff has rank zero,
[ A ^ ch ( R e f f ) ] 8 = − 1 6 T 4 − 1 6 p 1 c 2 ⟶ K − 1 3 ( λ ( T B ) + c ) c . (205) [\widehat{A}\operatorname{ch}(R_{\mathrm{eff}})]_8=-\frac{1}{6}T_4-\frac{1}{6}p_1c_2\overset{K}{\longrightarrow}-\frac{1}{3}(\lambda(TB)+c)c. \tag*{(205)} [ A ch ( R eff ) ] 8 = − 6 1 T 4 − 6 1 p 1 c 2 ⟶ K − 3 1 ( λ ( T B ) + c ) c . (205)
The restriction uses T 4 = 2 c 2 T_4 = 2c^2 T 4 = 2 c 2 and λ = p 1 / 2 \lambda= p_1/2 λ = p 1 /2 . Its tangent transgression is again − k 3 c / 3 -k_3c/3 − k 3 c /3 , proving that the total-space and base corrections match. Transporting the same differential trivialization gives h ∗ = h L C + k 3 h_* = h_{\mathrm{LC}} + k_3 h ∗ = h LC + k 3 , up to the retained exact-form and morphism data. Hence the disk term changes by + ∫ B k 3 c / 3 +\int_B k_3c/3 + ∫ B k 3 c /3 , while the closed eta term changes by its negative. Their corrections cancel in − a N − μ ∗ -a_N-\mu_* − a N − μ ∗ .
For this special connection λ ( A ∗ ) + c = 0 \lambda(A_*)+c=0 λ ( A ∗ ) + c = 0 as a form. The corresponding degree-four differential character is flat and lies in H 3 ( S 7 ; R / Z ) = 0 H^3(S^7;\mathbb{R}/\mathbb{Z})=0 H 3 ( S 7 ; R / Z ) = 0 , so it vanishes. The integral trivialization class is unique because H 3 ( S 7 ; Z ) = 0 H^3(S^7;\mathbb{Z})=0 H 3 ( S 7 ; Z ) = 0 , and its geometric representative can be chosen with h ∗ = 0 h_*=0 h ∗ = 0 . This recovers the selected zero local term used in Section 6, with the same finite connection transport. It is a property of that specified datum, not a vanishing statement for all sphere-bundle backgrounds.
A separate conditional boundary-family result
There is a different analytic question for the S 3 S^3 S 3 boundary fibres of N = D ( V ) → S 7 N=D(V)\to S^7 N = D ( V ) → S 7 . Suppose one chooses a smooth complex vertical Dirac-type realization of the difference anomaly whose Hermitian Clifford module, even-dimensional grading, twisting data, metrics and connections extend over the disk family and are product near its boundary. This assumption includes the identification of that realization with the complexified anomaly. Let D ∂ D_{\partial} D ∂ be its induced self-adjoint family on the closed boundary fibres. Its odd analytic index takes values in K 1 ( S 7 ) K^1(S^7) K 1 ( S 7 ) , the odd complex K K K -theory group of the base; this notation is distinct from the symmetry group K = Sp ( 1 ) K=\operatorname{Sp}(1) K = Sp ( 1 ) used above. Under these hypotheses, the odd analytic index of the boundary family vanishes [7 Appendix E, Proposition E.1] :
Ind C ( D ∂ ) = 0 ∈ K 1 ( S 7 ) . (206) \operatorname{Ind}_{\mathbb{C}}(D_{\partial})=0\in K^1(S^7). \tag*{(206)} Ind C ( D ∂ ) = 0 ∈ K 1 ( S 7 ) . (206)
The extending graded Clifford module supplies an even Dirac-symbol family on the compact disk fibres, and product structure identifies its induced boundary symbol over the whole base. This is precisely the situation of Melrose–Piazza’s family cobordism theorem [30 §§1–2, Proposition 3] . Carvalho gives the corresponding family-symbol extension formulation [31 ] . Cobordism invariance applies to an extending realization; it does not construct a realization of the anomaly.
For a smooth self-adjoint elliptic family on closed fibres over a compact base, vanishing of the odd analytic index is equivalent to existence of a Melrose–Piazza spectral section [30 §2, Proposition 1] . Applying this criterion to (D.27) therefore gives a spectral section for the assumed extending realization [7 Appendix E, Corollary E.2] . Such a section is a smooth family of self-adjoint pseudodifferential projections, equal to the identity on eigenspaces sufficiently far above a spectral cutoff and to zero sufficiently far below its negative. It is defined over the whole base, including points where eigenvalues cross zero. The theorem establishes existence; it selects neither a preferred section nor a spectral gap.
A characteristic-class check is available under further specified twisting hypotheses. For a spin-Dirac realization on M = Sp ( 2 ) → S 7 M=\operatorname{Sp}(2)\to S^7 M = Sp ( 2 ) → S 7 , distinguish the twisting bundle E \mathcal{E} E from the full Clifford module. If E \mathcal{E} E comes from an SU ( 2 ) \operatorname{SU}(2) SU ( 2 ) representation or is the complexification of a real bundle, its rational Chern character, like A ^ ( T π ) \hat{A}(T_\pi) A ^ ( T π ) for the vertical tangent bundle T π T_\pi T π , has components only in degrees divisible by four. The families index character is
ch ( Ind C ( D ∂ ) ) = ∫ M / S 7 A ^ ( T π ) ch ( E ) . (207) \operatorname{ch}\bigl(\operatorname{Ind}_{\mathbb{C}}(D_{\partial})\bigr)=\int_{M/S^{7}}\widehat{A}(T_{\pi})\operatorname{ch}(\mathcal{E}). \tag*{(207)} ch ( Ind C ( D ∂ ) ) = ∫ M / S 7 A ( T π ) ch ( E ) . (207)
Integration over three-dimensional fibres leaves positive odd degrees congruent to one modulo four. The only positive odd rational cohomology of S 7 S^{7} S 7 is in degree seven, so this character vanishes. Bott periodicity gives K 1 ( S 7 ) ≅ Z K^{1}(S^{7})\cong\mathbb{Z} K 1 ( S 7 ) ≅ Z ; rational injectivity and torsion-freeness then give integral vanishing. This check does not identify E \mathcal{E} E with the effective coefficient of the closed four-sphere calculation and does not replace the extending-operator hypotheses.
Neither a vanishing complex index nor an available complex spectral section supplies a real Pfaffian refinement. A boundary-family evaluation would still require a normalized taming, meaning a smoothing perturbation that makes the family invertible, together with its spectral-flow dependence, and compatible eta-form and real differential data. The closed reduction (D.22) and the direct APS calculation require none of those additional choices. Their returned observable remains (D.23), on the specified Spin, gauge and integral differential background.
Euler calibration and the two gauge embeddings
Two optional calculations illuminate different features of the response. An Euler integral fixes the sign and normalization of the effective seven-dimensional eta value in the Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) example of § 6 . A comparison of two gauge embeddings explains the cubic consistency of the response discussed in § 8 . The first calculation enlarges the base gauge bundle to a symplectic bundle on D 8 D^{8} D 8 ; the second keeps the eleven-dimensional disk and its complete boundary datum fixed. Neither is a premise of the direct twelve-dimensional computation in § 4 . The Euler calibration and the comparison of real gauge coefficients follow Li–Sun [7, Lemma C.1, Proposition C.2, Remark C.3; Appendix D].
The outward relative Euler unit
Let B = S ( H 2 ) = ∂ D 8 B=S(\mathbb{H}^{2})=\partial D^{8} B = S ( H 2 ) = ∂ D 8 have the outward orientation and let V y = y ⊥ H V_y=y^{\perp_{\mathbb{H}}} V y = y ⊥ H , with underlying complex rank-two bundle U U U given by right multiplication by i i i . Although this quaternionic line does not extend across the ball as a quaternionic line, its stabilization does: take the defining complex rank-four symplectic bundle W = D 8 × H 2 W=D^{8}\times\mathbb{H}^{2} W = D 8 × H 2 . On the boundary it splits as V ⊕ H V\oplus\mathbb{H} V ⊕ H , with unit section s ( y ) = y s(y)=y s ( y ) = y . Prescribe the connection a a a on V V V and the flat connection in the complementary quaternionic frame ( y , y i , y j , y k ) (y,yi,yj,yk) ( y , y i , y j , y k ) . The affine space of symplectic connections allows a smooth extension of this boundary connection.
This boundary frame specifies a relative Euler number, rather than an absolute characteristic class of the trivial bundle. For one quaternionic block, the complex orientation is opposite to the real orientation ( 1 , i , j , k ) (1,i,j,k) ( 1 , i , j , k ) ; for two blocks the signs multiply to + 1 +1 + 1 . Thus c 4 ( W ) = e ( W R ) c_{4}(W)=e(W_{\mathbb{R}}) c 4 ( W ) = e ( W R ) with the real orientation used here. The section s ( x ) = x s(x)=x s ( x ) = x has a single zero with derivative the identity. Its relative Euler number is consequently + 1 +1 + 1 .
The Atiyah–Patodi–Singer (APS) calculation needs the corresponding equality for the real Chern–Weil integral, not only an integral-cohomology statement. For two metric connections making s s s parallel on the boundary, choose a boundary frame beginning with s s s . Their curvature and connection variation have zero first row and column. The polarized Euler Pfaffian with one variation insertion therefore vanishes at the boundary. Chern–Weil transgression and Stokes show that their Euler integrals agree. A convenient comparison is the round hemisphere connection, transported to the trivial real bundle by its positive orthonormal frame. At the totally geodesic equator its outward normal is s ( y ) = y s(y)=y s ( y ) = y and is parallel. Its Euler density and volume give
∫ D 8 c 4 ( W ) = 105 16 π 4 16 π 4 105 = 1. (208) \int_{D^8} c_4(W)=\frac{105}{16\pi^4}\frac{16\pi^4}{105}=1. \tag*{(208)} ∫ D 8 c 4 ( W ) = 16 π 4 105 105 16 π 4 = 1. (208)
This comparison proves the integral for the prescribed symplectic connection. The hemisphere metric is used only to evaluate the Euler integral; the Dirac operators below use genuine product collars.
The effective eta value and finite tangent transport
The associated-family calculation supplies the effective complex coefficient R e f f = − 4 ⋅ 4 + 16 R_{\mathrm{eff}}=-4\cdot4+16 R eff = − 4 ⋅ 4 + 16 , where 4 4 4 is the defining representation of Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) , with its associated connection. Undecorated integers in complex coefficients denote trivial summands. On the boundary 4 ∣ Sp ( 1 ) = U + 2 4|_{\operatorname{Sp}(1)}=U+2 4 ∣ Sp ( 1 ) = U + 2 , so this coefficient restricts to − 4 U + 8 -4U+8 − 4 U + 8 . Appendix D identifies this representation and its associated connection with the effective base coefficient; a rank-zero replacement would lose the twisted base operator. Write ξ ( A ) = ( η A P S ( A ) + dim C ker A ) / 2 \xi(A)=(\eta_{\mathrm{APS}}(A)+\dim_{\mathbb{C}}\ker A)/2 ξ ( A ) = ( η APS ( A ) + dim C ker A ) /2 . For Levi–Civita tangent data and a product-collar spin metric on the ball, fix an inward coordinate u u u and the convention for the boundary Dirac operator A E A_E A E twisted by E E E :
D E + = c ( d u ) ( ∂ u + A E ) , ν = − ∂ u , c B ( v ) = c D ( ν ) c D ( v ) . (209) D_E^+=c(du)(\partial_u+A_E),\qquad\nu=-\partial_u,\qquad c_B(v)=c_D(\nu)c_D(v). \tag*{(209)} D E + = c ( d u ) ( ∂ u + A E ) , ν = − ∂ u , c B ( v ) = c D ( ν ) c D ( v ) . (209)
The APS condition projects out the nonnegative eigenspaces, including the zero eigenspace. Applying the twisted APS theorem to the actual symplectic twist and to the trivial complex line gives [8 Theorem (4.2), formula (4.3)]
ξ R L C = ∫ D 8 I ~ 8 − N R , ξ R = − 4 ξ ( A W ) + 16 ξ ( A 1 ) , (210) \xi_R^{\mathrm{LC}}=\int_{D^8}\widetilde I_8-N_R,\qquad\xi_R=-4\xi(A_W)+16\xi(A_1), \tag*{(210)} ξ R LC = ∫ D 8 I 8 − N R , ξ R = − 4 ξ ( A W ) + 16 ξ ( A 1 ) , (210)
N R = − 4 Ind A P S D W + + 16 Ind A P S D 1 + ∈ Z . N_R=-4\operatorname{Ind}_{\mathrm{APS}}D_W^++16\operatorname{Ind}_{\mathrm{APS}}D_1^+\in\mathbb{Z}. N R = − 4 Ind APS D W + + 16 Ind APS D 1 + ∈ Z .
The parallel trivial quaternionic line gives the operator splitting A W = A V ⊕ 2 A 1 A_W=A_V\oplus2A_1 A W = A V ⊕ 2 A 1 . Hence ξ R = − 4 ξ ( A V ) + 8 ξ ( A 1 ) = ξ e f f \xi_R=-4\xi(A_V)+8\xi(A_1)=\xi_{\mathrm{eff}} ξ R = − 4 ξ ( A V ) + 8 ξ ( A 1 ) = ξ eff , with all kernels included. The real normalization has already been converted into this integer combination of complex reduced eta values. There is no additional Majorana halving, and no congruence is divided by two. The integer N R N_R N R is retained until the final reduction modulo Z \mathbb{Z} Z ; it is not a substitute for the real parity argument in the twelve-dimensional calculation.
Put Z 4 = λ + c Z_4=\lambda+c Z 4 = λ + c and W 4 ′ = − c / 3 W'_4=-c/3 W 4 ′ = − c /3 , with λ = p 1 / 2 \lambda=p_1/2 λ = p 1 /2 . On D 8 D^8 D 8 , c c c denotes c 2 ( W ) c_2(W) c 2 ( W ) ; its restriction is the original c = c 2 ( U ) c=c_2(U) c = c 2 ( U ) on B B B . For roots x , − x , y , − y x,-x,y,-y x , − x , y , − y , the degree-four and degree-eight components of ch ( W ) \operatorname{ch}(W) ch ( W ) are − c -c − c and T 4 / 24 T_4/24 T 4 /24 . The rank of R e f f R_{\mathrm{eff}} R eff is zero. Combining its degree-eight character with the product of A ^ 4 = − λ / 12 \widehat{A}_4=-\lambda/12 A 4 = − λ /12 and its degree-four character gives the full effective eight-form:
T 4 = 2 ( x 4 + y 4 ) , I ~ 8 = − 1 6 T 4 − 1 3 λ c , T 4 = 2 c 2 − 4 c 4 ( W ) . (211) T_4=2(x^4+y^4),\qquad\widetilde{I}_8=-\frac{1}{6}T_4-\frac{1}{3}\lambda c,\qquad T_4=2c^2-4c_4(W). \tag*{(211)} T 4 = 2 ( x 4 + y 4 ) , I 8 = − 6 1 T 4 − 3 1 λ c , T 4 = 2 c 2 − 4 c 4 ( W ) . (211)
Since H 4 ( D 8 ; R ) = 0 H^4(D^8;\mathbb{R})=0 H 4 ( D 8 ; R ) = 0 , the extended form Z 4 Z_4 Z 4 has a primitive. Since H 3 ( S 7 ; R ) = 0 H^3(S^7;\mathbb{R})=0 H 3 ( S 7 ; R ) = 0 , that primitive can be adjusted by an exact term to agree with the specified curving h h h on the boundary. Subtracting the boundary term by Stokes cancels the entire tangent-dependent part of the density:
I ~ 8 − Z 4 W 4 ′ = − 1 6 T 4 + 1 3 c 2 = 2 3 c 4 ( W ) , (212) \widetilde{I}_8-Z_4W'_4=-\frac{1}{6}T_4+\frac{1}{3}c^2=\frac{2}{3}c_4(W), \tag*{(212)} I 8 − Z 4 W 4 ′ = − 6 1 T 4 + 3 1 c 2 = 3 2 c 4 ( W ) , (212)
ξ e f f L C ( B ) − ∫ B h W 4 ′ = 2 3 ( m o d Z ) . (213) \xi_{\mathrm{eff}}^{\mathrm{LC}}(B)-\int_B hW'_4=\frac{2}{3}\pmod{\mathbb{Z}}. \tag*{(213)} ξ eff LC ( B ) − ∫ B h W 4 ′ = 3 2 ( mod Z ) . (213)
The residual polynomial vanishes on the boundary Sp ( 1 ) \operatorname{Sp}(1) Sp ( 1 ) reduction, and its real integral is fixed by (E.1).
For a specified stable tangent connection A ∗ A_* A ∗ , keep the gauge connection fixed and set k 3 = T λ ( A L C , A ∗ ) k_3=T_\lambda(A_{\mathrm{LC}},A_*) k 3 = T λ ( A LC , A ∗ ) . The tangent dependence − λ c / 3 -\lambda c/3 − λ c /3 determines the finite correction. Integrating Chern–Weil variation along the connection path [24 Proposition 3.8] , extended over the same disk with product collars, changes the density integral by − ∫ B k 3 c / 3 -\int_B k_3c/3 − ∫ B k 3 c /3 . The corrected differential-index value therefore retains the same integer N R N_R N R in (E.3). In the following formula, b b b denotes a global real two-form encoding the exact change of curving, distinct from the integral trivializing cochain in Appendix § 13 :
ξ e f f ∗ = ξ e f f L C − 1 3 ∫ B k 3 c , h ∗ = h L C + k 3 + d b . (214) \xi_{\mathrm{eff}}^*=\xi_{\mathrm{eff}}^{\mathrm{LC}}-\frac{1}{3}\int_B k_3c,\qquad h_*=h_{\mathrm{LC}}+k_3+db. \tag*{(214)} ξ eff ∗ = ξ eff LC − 3 1 ∫ B k 3 c , h ∗ = h LC + k 3 + d b . (214)
The change in ∫ B h W 4 ′ \int_B hW'_4 ∫ B h W 4 ′ is the same − ∫ B k 3 c / 3 -\int_B k_3c/3 − ∫ B k 3 c /3 , since c c c is closed. Thus the corrected eta-minus-local combination remains 2 / 3 2/3 2/3 . This uses ordinary Levi–Civita APS operators and a differential-index correction, not a raw spectrum for the stable tangent connection. A change of affine comparison path changes the transgression by an exact form through the connection-simplex identity; its integral against closed c c c vanishes. For the selected Sp ( 2 ) \operatorname{Sp}(2) Sp ( 2 ) representative with h ∗ = 0 h_*=0 h ∗ = 0 , it gives μ ∗ = 2 / 3 \mu_*=2/3 μ ∗ = 2/3 ; the response convention ϕ = − a N − μ ∗ \phi=-a_N-\mu_* ϕ = − a N − μ ∗ then returns ϕ = 1 / 3 \phi=1/3 ϕ = 1/3 and O = e 2 π i / 3 \mathcal{O}=e^{2\pi i/3} O = e 2 π i /3 as in § 6 .
The outward Euler unit must also be distinguished from the published clutching generator. Tachikawa–Zhang normalize that generator by ∫ S 8 T 4 / 48 = 1 \int_{S^8}T_4/48=1 ∫ S 8 T 4 /48 = 1 and use the homotopy map with multiplier 12 [3 equations (42), (54)–(55)] . Degree-four products integrate to zero on S 8 S^8 S 8 , so this convention gives ∫ S 8 c 4 = − 12 \int_{S^8}c_4=-12 ∫ S 8 c 4 = − 12 . With the cap extending the ball and its boundary section by a trivial cap, it is − 12 -12 − 12 times the geometric relative unit above. The relative homotopy generator whose twelvefold multiple is this clutching class therefore has Euler number − 1 -1 − 1 . For R e f f R_{\mathrm{eff}} R eff the primitive coefficient is 48 ( − 1 / 6 ) = − 8 48(-1/6)=-8 48 ( − 1/6 ) = − 8 , giving the value − 8 / 12 = − 2 / 3 -8/12=-2/3 − 8/12 = − 2/3 on the source generator. Our outward relative unit is its negative, so the additive eta-minus-local value changes sign and becomes + 2 / 3 +2/3 + 2/3 . The two numbers refer to opposite generators; the direct Euler integral fixes the value here without using the homotopy normalization.
Two arrangements of the same boundary gauge datum
The embedding comparison applies to the general sphere-bundle background of § 3 : a quaternionic line V V V over a closed connected spin seven-manifold B B B , its principal Sp ( 1 ) \operatorname{Sp}(1) Sp ( 1 ) bundle P P P , and the specified integral differential trivialization of λ ( T B ) + c \lambda(TB)+c λ ( T B ) + c . No rational-cohomology vanishing is required for this comparison. On N = D ( V ) N=D(V) N = D ( V ) , keep the tangent connection, quaternionic connection and integral differential datum H ˇ N \check{H}_N H ˇ N fixed. The arrangement used in the main calculation is E 20 = 2 Q H ⊕ R 8 E_{20}=2Q_{\mathrm{H}}\oplus\mathbb{R}^{8} E 20 = 2 Q H ⊕ R 8 , E 20 ′ = R 16 E'_{20}=\mathbb{R}^{16} E 20 ′ = R 16 , where Q = π ∗ P Q=\pi^*P Q = π ∗ P and Q H Q_{\mathrm{H}} Q H is its associated real quaternionic line. Write C = c 2 ( Q , A ) C=c_2(Q,A) C = c 2 ( Q , A ) at the retained connection A A A . Compare it with
E 11 = Q H ⊕ R 12 , E 11 ′ = Q H ⊕ R 12 . (215) E_{11}=Q_{\mathrm{H}}\oplus\mathbb{R}^{12}, \qquad E'_{11}=Q_{\mathrm{H}}\oplus\mathbb{R}^{12}. \tag*{(215)} E 11 = Q H ⊕ R 12 , E 11 ′ = Q H ⊕ R 12 . (215)
The labels count quaternionic copies in the two ordered factors; they are not fluxes or current-algebra levels. Both constructions give actual Spin ( 16 ) 2 \operatorname{Spin}(16)^2 Spin ( 16 ) 2 bundles with total gauge class − 2 C -2C − 2 C .
Equality of this class does not by itself identify the boundary response. Order the total real blocks as ( Q 1 , Q 2 , R 8 , R 16 ) (Q_1,Q_2,\mathbb{R}^{8},\mathbb{R}^{16}) ( Q 1 , Q 2 , R 8 , R 16 ) in the first arrangement and ( Q 1 , R 12 , Q 2 , R 12 ) (Q_1,\mathbb{R}^{12},Q_2,\mathbb{R}^{12}) ( Q 1 , R 12 , Q 2 , R 12 ) in the second. Moving the second four-plane across twelve trivial directions has determinant ( − 1 ) 4 ⋅ 12 = 1 (-1)^{4\cdot12}=1 ( − 1 ) 4 ⋅ 12 = 1 . Fix its constant Spin ( 32 ) \operatorname{Spin}(32) Spin ( 32 ) lift. This intertwines the total connections and the combined tautological frames. It identifies the additive gauge data, while keeping the two anomaly coefficient systems distinct.
Retain the quaternionic and Whitney characteristic morphisms of Appendix § 13 , in the order ( Q 1 , Q 2 ) (Q_1,Q_2) ( Q 1 , Q 2 ) . For q ˇ = c ˇ 2 ( Q , A ) \check{q}=\check{c}_2(Q,A) q ˇ = c ˇ 2 ( Q , A ) , both arrangements then have total gauge cocycle ( − q ˇ ) + ( − q ˇ ) (-\check{q})+(-\check{q}) ( − q ˇ ) + ( − q ˇ ) and boundary trivialization − 2 γ ˇ τ -2\check{\gamma}_{\tau} − 2 γ ˇ τ , including their integral and real cochains. The same H ˇ N \check{H}_N H ˇ N therefore induces the same s ˇ M = H ˇ N ∣ M − 2 γ ˇ τ \check{s}_M=\check{H}_N|_M-2\check{\gamma}_{\tau} s ˇ M = H ˇ N ∣ M − 2 γ ˇ τ . No independent degree-three class is inserted. The tangent Spin \operatorname{Spin} Spin structure, normal splitting, connection and product collars also agree. Trivialize each boundary Spin ( 16 ) \operatorname{Spin}(16) Spin ( 16 ) factor by its own specified frame, preserving the factor order. Each actual boundary coefficient is then 256 − 128 − 128 = 0 256-128-128=0 256 − 128 − 128 = 0 as a virtual real representation. Their normalized reference elements coincide, without a square-root or Pfaffian-sign choice. Independently chosen nullhomotopies with equal curvature would not suffice.
The full invariant polynomials distinguish the two interiors. Since p ( Q H ) = 1 − 2 C + C 2 p(Q_{\mathrm{H}})=1-2C+C^2 p ( Q H ) = 1 − 2 C + C 2 , the first arrangement has ( n , n ′ ) = ( − 2 C , 0 ) (n,n')=(-2C,0) ( n , n ′ ) = ( − 2 C , 0 ) and ( p 2 ( E ) , p 2 ( E ′ ) ) = ( 6 C 2 , 0 ) (p_2(E),p_2(E'))=(6C^2,0) ( p 2 ( E ) , p 2 ( E ′ )) = ( 6 C 2 , 0 ) . The second has ( n , n ′ ) = ( − C , − C ) (n,n')=(-C,-C) ( n , n ′ ) = ( − C , − C ) and ( p 2 ( E ) , p 2 ( E ′ ) ) = ( C 2 , C 2 ) (p_2(E),p_2(E'))=(C^2,C^2) ( p 2 ( E ) , p 2 ( E ′ )) = ( C 2 , C 2 ) . Substitution into the full Y 8 Y_8 Y 8 of (7) gives
Y 20 = − 1 3 C 2 , Y 11 = 1 6 C 2 , Y 20 = − 2 Y 11 . (216) Y_{20}=-\frac{1}{3}C^2,\qquad Y_{11}=\frac{1}{6}C^2,\qquad Y_{20}=-2Y_{11}. \tag*{(216)} Y 20 = − 3 1 C 2 , Y 11 = 6 1 C 2 , Y 20 = − 2 Y 11 . (216)
The corresponding spectral relation requires an identity of real coefficients. In each arrangement the difference coefficient is the tensor product of the two vector representations minus the two positive real half-spin representations of § 2 . If u u u is the character of U U U , the one-copy vector and half-spin characters are 2 u + 12 2u+12 2 u + 12 and 32 ( u + 2 ) 32(u+2) 32 ( u + 2 ) ; the two-copy characters are 4 u + 8 4u+8 4 u + 8 and 8 ( u + 2 ) 2 8(u+2)^2 8 ( u + 2 ) 2 . Thus the two ordered subtractions have characters
16 ( 4 u + 8 ) − 8 ( u + 2 ) 2 − 128 = − 8 u 2 + 32 u − 32 , 16(4u+8)-8(u+2)^2-128=-8u^2+32u-32, 16 ( 4 u + 8 ) − 8 ( u + 2 ) 2 − 128 = − 8 u 2 + 32 u − 32 ,
( 2 u + 12 ) 2 − 64 ( u + 2 ) = 4 u 2 − 16 u + 16. (2u+12)^2-64(u+2)=4u^2-16u+16. ( 2 u + 12 ) 2 − 64 ( u + 2 ) = 4 u 2 − 16 u + 16.
Using Ad ( U ) = Sym 2 U \operatorname{Ad}(U) = \operatorname{Sym}^{2} U Ad ( U ) = Sym 2 U , with character u 2 − 1 u^{2}-1 u 2 − 1 , their ordered differences give
Δ 20 , C = 32 U − 8 Ad ( U ) − 40 , Δ 11 , C = − 16 U + 4 Ad ( U ) + 20 , Δ 20 , R = 16 H − 8 Ad R − 40 R , Δ 11 , R = − 8 H + 4 Ad R + 20 R . (217) \begin{aligned}
\Delta_{20,\mathbb{C}} &= 32U - 8\operatorname{Ad}(U) - 40, \qquad&\Delta_{11,\mathbb{C}} &= -16U + 4\operatorname{Ad}(U) + 20,\\
\Delta_{20,\mathbb{R}} &= 16\mathbb{H} - 8\operatorname{Ad}_{\mathbb{R}} - 40\mathbb{R}, \qquad&\Delta_{11,\mathbb{R}} &= -8\mathbb{H} + 4\operatorname{Ad}_{\mathbb{R}} + 20\mathbb{R}. \tag*{(217)}
\end{aligned} Δ 20 , C Δ 20 , R = 32 U − 8 Ad ( U ) − 40 , = 16 H − 8 Ad R − 40 R , Δ 11 , C Δ 11 , R = − 16 U + 4 Ad ( U ) + 20 , = − 8 H + 4 Ad R + 20 R . (217)
Here H C = U ⊕ U \mathbb{H}_{\mathbb{C}} = U \oplus U H C = U ⊕ U ; the adjoint and trivial representations are real. The real identity is also recovered from H ⊗ R H = 4 Ad R + 4 R \mathbb{H} \otimes_{\mathbb{R}} \mathbb{H} = 4\operatorname{Ad}_{\mathbb{R}} + 4\mathbb{R} H ⊗ R H = 4 Ad R + 4 R . Orthogonal compact-group intertwiners become parallel after association, including the clutching. Thus Δ 20 , R = − 2 Δ 11 , R \Delta_{20,\mathbb{R}} = -2\Delta_{11,\mathbb{R}} Δ 20 , R = − 2 Δ 11 , R relates the actual virtual real closed Dirac systems, with kernels and the same finite connection corrections. Equality of ranks, or equality of local polynomials alone, would not establish this spectral statement.
Additivity now yields a 20 = − 2 a 11 a_{20}=-2a_{11} a 20 = − 2 a 11 , μ 20 = − 2 μ 11 \mu_{20}=-2\mu_{11} μ 20 = − 2 μ 11 and ϕ 20 = − 2 ϕ 11 \phi_{20}=-2\phi_{11} ϕ 20 = − 2 ϕ 11 modulo Z \mathbb{Z} Z . Filling independence for the synchronized complete boundary datum supplies the separate equality ϕ 20 = ϕ 11 \phi_{20}=\phi_{11} ϕ 20 = ϕ 11 . Combining them gives
3 ϕ = 0 in R / Z , O 3 = 1. (218) 3\phi=0\quad\text{in }\mathbb{R}/\mathbb{Z},\qquad O^{3}=1. \tag*{(218)} 3 ϕ = 0 in R / Z , O 3 = 1. (218)
This argument uses the normalized filling independence stated in § 2 , including the real gluing application discussed in Appendix § 14 . The characteristic-class comparison does not discharge that boundary-line identification.
For comparison, the associated-family reduction of the ( 1 , 1 ) (1,1) ( 1 , 1 ) system has effective coefficient 2 U − 4 2U-4 2 U − 4 and local term − ∫ B h c / 6 -\int_{B}hc/6 − ∫ B h c /6 , and gives ϕ 11 = − K / 6 \phi_{11}=-K/6 ϕ 11 = − K /6 [7 Appendix D] . The difference from the main value ϕ 20 = K / 3 \phi_{20}=K/3 ϕ 20 = K /3 is ϕ 20 − ϕ 11 = K / 2 \phi_{20}-\phi_{11}=K/2 ϕ 20 − ϕ 11 = K /2 , which is integral because K K K is even. On exactly the domain H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 H^{3}(B;\mathbb{Q})=H^{4}(B;\mathbb{Q})=0 H 3 ( B ; Q ) = H 4 ( B ; Q ) = 0 , where the Crowley–Goette invariant t t t is defined as in § 5 , this is − 4 t = 8 t -4t=8t − 4 t = 8 t , with 12 t = K / 2 = 0 12t=K/2=0 12 t = K /2 = 0 in R / Z \mathbb{R}/\mathbb{Z} R / Z . These relations preserve the actual coefficients; they do not divide an eta congruence by two. The embedding comparison explains cubic consistency, while the relative pairing in the main calculation determines which of the three response values occurs. Its reduced six-dimensional meaning remains the anomaly response of § 8 , without a vacuum-counting interpretation.