Polarization-Resolved MDCS: Advances in Coherent Spectroscopy

• Polarization-resolved MDCS is a nonlinear spectroscopy technique that uses controlled excitation and detection polarizations to isolate specific quantum pathways and transition symmetries.
• The method employs phase-stable pulse sequences and Fourier-transform analysis to separate homogeneous and inhomogeneous broadening, enabling detailed insights into vibronic and many-body correlations.
• It has been applied to diverse systems such as semiconductor microcavities, cuprate superconductors, and π-conjugated polymers to resolve Raman, excitonic, and spin coherence dynamics.

Polarization-resolved multidimensional coherent spectroscopy (MDCS) denotes a class of nonlinear coherent spectroscopies in which the phase-stable pulse sequence of MDCS is combined with controlled excitation and detection polarizations so that selected Liouville pathways, transition symmetries, or operator components dominate the measured multidimensional response. In the standard optical implementation, the experiment measures a third-order four-wave-mixing signal and Fourier transforms its delay-dependent amplitude and phase to obtain spectra that correlate excitation and emission energies; polarization control then adds symmetry selectivity to the usual MDCS capabilities of separating homogeneous and inhomogeneous broadening, resolving cross-peaks, and detecting vibronic and many-body correlations (Gutiérrez-Meza et al., 2022). Across current applications, polarization selection has been used to isolate Raman coherences of ground-state spins, suppress biexciton-like channels in semiconductor microcavities, select $B_{1g}$ or $B_{2g}$ Raman symmetries in cuprates, and distinguish alternating and uniform probe couplings in candidate fractionalized pyrochlores (Salewski et al., 2017, Mishra et al., 23 Sep 2025, Potts et al., 2023).

1. Nonlinear-response framework

In third-order MDCS, the induced polarization is expanded as

$\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$

For isotropic media, the second-order response vanishes, so the dominant coherent signal is usually third order (Gutiérrez-Meza et al., 2022). In the standard noncollinear BOXCARS geometry, three excitation pulse trains are focused onto the sample and the four-wave-mixing signal is emitted in a phase-matched direction, while a local oscillator co-propagates with the signal for heterodyne detection and spectral interferometry. The two common phase-matching directions are the rephasing channel,

$\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$

and the nonrephasing channel,

$\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$

with the total nonlinear response given by the sum of rephasing and nonrephasing spectra (Gutiérrez-Meza et al., 2022).

The basic pulse ordering is likewise standard. Pulse 1 creates a ground-state/excited-state coherence during the first delay, pulse 2 converts that coherence into a population or another coherence during the waiting time, and pulse 3 creates a new coherence that radiates the detected signal. The resulting 2D map is constructed by Fourier transforming along the first coherence time to obtain the excitation-energy axis, while the emission-energy axis is obtained from spectral detection of the emitted field. Because the signal is measured heterodynely with the local oscillator, both amplitude and phase are retained (Gutiérrez-Meza et al., 2022). In double-quantum MDCS, the phase-conjugated pulse arrives last, so the system evolves first through a double-quantum coherence before emission; the relevant spectral axes then correlate the double-quantum evolution frequency with the emission frequency (Lomsadze et al., 2020).

Within this framework, polarization enters through the tensor character of the nonlinear susceptibility. Even when a study does not develop a full tensor decomposition, the choice of pulse polarizations determines which pathways survive and which couplings are emphasized. This is explicit in DQ-MDCS simulations that assume co-polarized pulses to simplify the analysis, so that peak shape can be attributed primarily to resonance correlations, interaction-induced shifts, and the balance of homogeneous and inhomogeneous broadening (Lomsadze et al., 2020).

2. Polarization as a pathway- and symmetry-selection variable

The general role of polarization control in MDCS is pathway selection. In $\pi$-conjugated polymers, polarization-resolved MDCS is described as a means to select specific Liouville pathways, enhance or suppress certain transition symmetries, help distinguish coupling versus orientational averaging, and isolate vibronic or electronic contributions in anisotropic films. The tensor nature of the response is already encoded in $\chi^{(3)}$, so polarization is an available control knob even when the primary discussion emphasizes phase matching and temporal sequence rather than explicit polarization combinations (Gutiérrez-Meza et al., 2022).

In semiconductor spin spectroscopy, polarization selection can qualitatively change the dominant pathway. In the CdTe/(Cd,Mg)Te quantum-well experiment, the co-polarized HHH channel mainly converts optical coherence into populations, whereas the cross-polarized HVV channel drives a stimulated step-like Raman process that transfers coherence into the ground-state electron-spin manifold. In that configuration, the emitted signal acquires sidebands at $\pm\omega_L$ relative to the optical carrier, and Fourier transformation along the Raman-evolution interval yields peaks at $\Omega_2=\pm\omega_L$, with widths set by the electron transverse spin dephasing rate (Salewski et al., 2017).

In semiconductor microcavities, polarization is used to simplify the many-body level scheme. Co-circular $\sigma^+\sigma^+\sigma^+\sigma^+$ excitation suppresses biexciton-like contributions and keeps the interpretation focused on lower- and upper-polariton dynamics, especially the self- and mutual-interaction features (Paul et al., 2021). Conversely, in the detuning-dependent microcavity study, a biexcitonic companion feature appears only in collinear polarization and is suppressed for co-circular polarization, consistent with biexciton selection rules (Wilmer et al., 2015).

In strongly correlated quantum materials, polarization can act as a direct symmetry selector rather than merely a pathway filter. In underdoped Bi-2212, cross-linear pulse configurations together with crystallographic alignment select either $B_{2g}$0 symmetry along the Cu–Cu direction or $B_{2g}$1 symmetry along the Cu–O direction, thereby mapping predominantly onto antinodal and nodal regions of the Fermi surface, respectively (Mishra et al., 23 Sep 2025). In Ce$B_{2g}$2Zr$B_{2g}$3O$B_{2g}$4 and Nd$B_{2g}$5Zr$B_{2g}$6O$B_{2g}$7, changing the probe-field direction changes the magnetization operator and therefore the matrix elements entering the nonlinear response: $B_{2g}$8 couples to the $B_{2g}$9 chains, $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$0 to the $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$1 chains, and $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$2 couples to both but with a different momentum structure (Potts et al., 2023).

3. Lineshapes, cross-peaks, and correlation under polarization control

A central advantage of MDCS over linear spectroscopy is the separation of homogeneous and inhomogeneous broadening. In a rephasing experiment, diagonal and cross-diagonal widths carry different information: if the inhomogeneous width $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$3 is much larger than the homogeneous dephasing rate $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$4, then the diagonal cut is Gaussian with width $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$5, while the cross-diagonal cut is Lorentzian-like with width $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$6; in the opposite limit, both cuts are Lorentzian with width $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$7 (Gutiérrez-Meza et al., 2022). This distinction is operationally important in systems whose spectra are dominated by disorder, vibronic structure, or static variability.

Cross-peaks carry equally direct interpretive weight. Diagonal peaks correspond to autocorrelations of optical transitions, while cross-peaks arise when two resonances are correlated, often through a common ground state or through coupling between excited states. Off-diagonal structure is therefore not simply disorder; it is evidence of real physical coupling, and the intensity and shape of a cross-peak report on coupling strength and relaxation between states (Gutiérrez-Meza et al., 2022). In polarization-resolved experiments this is often decisive, because symmetry selection can suppress competing diagonal pathways and expose otherwise weak inter-state couplings.

Double-quantum MDCS makes the correlation content of the lineshape especially explicit. When correlated inhomogeneous broadening is included through a 2D Gaussian distribution with correlation coefficient $\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$8, the DQ peak ellipticity

$\Vec{P}(t) = \epsilon_0 \bigl( \chi^{(1)} \Vec{E}_A(t) + \chi^{(2)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t) + \chi^{(3)} \Vec{E}_A(t)\!\cdot\!\Vec{E}_B(t)\!\cdot\!\Vec{E}_C(t) + \ldots \bigr).$9

tracks the resonance correlation: $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$0 gives ellipticity $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$1, $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$2 gives ellipticity $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$3, and $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$4 gives ellipticity $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$5 (Lomsadze et al., 2020). Diagonal elongation is thus a proxy for correlation between emission and double-quantum frequencies. The same work also stresses an important caveat: peak orientation is not controlled by correlation alone, because strong homogeneous broadening or extra decoherence can tilt peaks vertically and mask the intrinsic diagonal elongation (Lomsadze et al., 2020).

Zero-quantum spectra provide an analogous logic for low-energy Raman coherences. In Bi-2212, a discrete Raman mode would produce a narrow, horizontally elongated feature, a broad uncorrelated Raman continuum would produce a broad peak, and a correlated or anti-correlated Raman-electronic response would appear as a diagonal or cross-diagonal structure. In that representation, the cross-diagonal width becomes the key observable because the 0Q lineshape depends on the dephasing of the Raman coherence, the dephasing of the optical electronic coherence, and the strength of the correlation between them (Mishra et al., 23 Sep 2025).

4. Experimental implementations and readout architectures

The canonical architecture for polarization-resolved MDCS remains heterodyne-detected four-wave mixing in a noncollinear geometry, but the literature also establishes collinear and comb-based implementations in which polarization selection can be layered onto alternative readout schemes. In the photocurrent-detected MD-COPS method, four pulses $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$6, $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$7, $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$8, and $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_B-\Vec{k}_A+\Vec{k}_C,$9 arrive collinearly, each tagged by a distinct acousto-optical modulation frequency, and the desired four-wave-mixing pathways are isolated as radio-frequency beat notes rather than by optical phase matching. The auxiliary continuous-wave laser provides reference beat notes for lock-in detection, partial compensation of mechanical fluctuations, and direct acquisition of the complex signal $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$0, which is then Fourier transformed to construct the multidimensional spectrum (Nardin et al., 2013).

That scheme preserves essential MDCS features—phase sensitivity, multidimensional Fourier analysis, and separation of rephasing and non-rephasing pathways—while removing the requirement of a far-field wave-vector-selected beam. It is therefore applicable to electrically contacted nanostructures that do not emit a well-defined four-wave-mixing beam, including single quantum dots, nanowires, and carbon nanotubes (Nardin et al., 2013). The paper is explicit, however, that polarization resolution is not a central element of the demonstrated technique, so the method is best understood as a detection architecture that can be combined with, but does not itself define, polarization-resolved MDCS (Nardin et al., 2013).

Tri-comb spectroscopy provides a different implementation strategy. It uses three optical frequency combs with slightly different repetition rates, one photodetector, and no mechanical moving elements. After digitization, the multidimensional coherent spectrum is constructed by a two-dimensional Fourier transform with respect to rotated time coordinates, and the demonstration on Doppler-broadened rubidium produced a multidimensional coherent spectrum with comb cross-diagonal resolution from only $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$1 ms of data (Lomsadze et al., 2018). The experiment employed an HVVH configuration, with Comb 1 horizontally polarized and Comb 2 vertically polarized, showing that polarization-controlled excitation is compatible with comb-based MDCS even though the paper does not develop a full polarization-resolved tensor analysis (Lomsadze et al., 2018).

These implementations indicate that polarization resolution and signal acquisition are separable design dimensions. A box-geometry experiment, a collinear frequency-tagged experiment, and a tri-comb experiment can all implement the same basic delay-space and Fourier-space logic; what changes is how the nonlinear field is isolated and phase referenced.

5. Representative physical systems and observables

In $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$2-conjugated polymers and related blends, MDCS has been used to reveal excitonic structure and dynamics hidden in linear absorption and photoluminescence. Rephasing experiments on PBTTT showed a suppressed $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$3 diagonal peak relative to $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$4, cross-peaks between vibronic resonances, unexpected excited-state absorption structure in the $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$5 cross peak, and nearly symmetric diagonal peak shapes indicating moderate inhomogeneity and significant homogeneous broadening. In P3HT, MDCS showed a homogeneous linewidth much larger than that of isolated chains and dephasing that could not be explained by a simple distribution of noninteracting single-chain spectra. Population-time-resolved measurements on P3HT and P3HT:PCBM revealed vibrational coherences at $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$6 cm$\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$7, different dephasing times for ground- and excited-state coherences, and ultrafast electron transfer in blends, while two-quantum coherence measurements in PBTTT directly resolved Frenkel biexcitons with attractive H-like and repulsive J-like correlations (Gutiérrez-Meza et al., 2022).

In semiconductor microcavities, polarization-resolved MDCS has been used both to control the accessible many-body channels and to interpret waiting-time dynamics. In one study, co-circular excitation suppressed biexciton-like contributions and enabled a cleaner assignment of lower- and upper-polariton self- and mutual-interaction features; fast decay components, oscillations in the off-diagonal peaks at the LP–UP splitting, and zero-quantum features at nonzero $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$8 established the presence of excited-state coherence rather than simple relaxation (Paul et al., 2021). In another, rephasing and non-rephasing spectra mapped the cavity–exciton anti-crossing, resolved diagonal intra-action peaks and off-diagonal interaction peaks, and identified a biexcitonic feature shifted from the exciton by approximately $\Vec{k}_{S_{\mathrm{FWM}}}=\Vec{k}_A-\Vec{k}_B+\Vec{k}_C,$9 meV, with detuning-dependent amplitude inversion of the cross-peaks as the lower polariton crossed the biexciton energy (Wilmer et al., 2015).

In localized semiconductor spin systems, polarization-resolved MDCS converts a tiny ground-state Zeeman splitting into an optically measurable Raman-coherence feature. In the CdTe/(Cd,Mg)Te quantum well, Raman-coherence spectra in the cross-polarized channel exhibited peaks at $\pi$0, whose widths yielded the electron transverse spin dephasing rate and whose positions yielded the Zeeman splitting $\pi$1. The approach resolved sub-$\pi$2eV differences between donor-bound electrons and electrons localized at potential fluctuations even though the homogeneous linewidth of the optical transitions was larger by two orders of magnitude (Salewski et al., 2017).

In underdoped Bi-2212, polarization resolution provides direct symmetry selection between nodal and antinodal Raman dynamics. In the superconducting phase at $\pi$3 K, the $\pi$4 nodal response showed a cross-diagonal linewidth of about $\pi$5 meV, corresponding to a decay time of $\pi$6 fs, together with an anti-correlation between the Raman energy $\pi$7 and the optical excitation energy $\pi$8 eV. The $\pi$9 antinodal response decohered in $\chi^{(3)}$0 fs and showed no measurable correlations. The narrow diagonal structure was observed only in the superconducting phase and disappeared in the pseudogap and normal phases (Mishra et al., 23 Sep 2025).

In the candidate fractionalized pyrochlores Ce$\chi^{(3)}$1Zr$\chi^{(3)}$2O$\chi^{(3)}$3 and Nd$\chi^{(3)}$4Zr$\chi^{(3)}$5O$\chi^{(3)}$6, polarization dependence is used to select microscopic operators in the nonlinear magnetic response. The rephasing or spinon-echo component appears as a streak along $\chi^{(3)}$7, and the polarization dependence distinguishes spinon continua on the $\chi^{(3)}$8 chains from magnon or two-magnon responses on the field-polarized $\chi^{(3)}$9 chains. The $\pm\omega_L$0 probe is particularly sensitive to the lower band edge of the spinon continuum in Ce$\pm\omega_L$1Zr$\pm\omega_L$2O$\pm\omega_L$3, while the relative intensity of one- and two-magnon signals in Nd$\pm\omega_L$4Zr$\pm\omega_L$5O$\pm\omega_L$6 probes the dipolar-octupolar mixing angle $\pm\omega_L$7 (Potts et al., 2023).

6. Conceptual boundaries, recurrent misconceptions, and current scope

A recurrent methodological point is that polarization-controlled MDCS is not automatically equivalent to a fully polarization-resolved tensor analysis. Several papers explicitly use fixed polarization configurations—co-polarized pulses in DQ-MDCS, HVVH in tri-comb spectroscopy, or polarization-controlled excitation in box or collinear geometries—without deriving full $\pm\omega_L$8 decompositions or comparing multiple polarization channels systematically (Lomsadze et al., 2020, Lomsadze et al., 2018). Conversely, some studies are explicitly about polarization dependence in the deeper sense that changing the field direction changes the operator being probed and therefore the excitation channel itself, as in the pyrochlore and cuprate examples (Potts et al., 2023, Mishra et al., 23 Sep 2025).

The literature also corrects several interpretive simplifications. A continuum in 2DCS is not automatically evidence for fractionalization, because in Ce$\pm\omega_L$9Zr$\Omega_2=\pm\omega_L$0O$\Omega_2=\pm\omega_L$1 the $\Omega_2=\pm\omega_L$2-chain response can look streak-like due to conventional two-magnon continua; it is the combination of polarization dependence and rephasing structure that separates spinon continua from multi-magnon continua (Potts et al., 2023). Likewise, a nonzero double-quantum signal implies interaction-induced symmetry breaking, because without interactions the positive and negative pathways cancel, but the observed DQ peak shape still depends strongly on dephasing and not on correlation alone (Lomsadze et al., 2020). In polymer spectroscopy, off-diagonal structure is not simply disorder but evidence of real physical coupling (Gutiérrez-Meza et al., 2022).

Recent theory papers extend MDCS into molecular polaritons, correlated lattice models, and light-driven superconductors, but several of those works also note that their present formulations are not polarization-resolved in the strict sense. Open-system Liouvillian theory for molecular polaritons states that polarization-dependent dipoles and selected polarization combinations could be incorporated by replacing scalar dipoles with vector or tensor dipoles and recalculating the excitation and detection masks (Gallego-Valencia et al., 2024). A nonequilibrium DMFT formulation for correlated lattice systems explicitly states that it focuses on a collinear setup that does not exploit the polarization of light, while suggesting that additional information could be obtained by varying geometry, polarization vectors, and relative phases (Chen et al., 2024). A THz-MDCS theory for multi-band superconductors emphasizes phase-locked collinear fields and gauge-invariant transport, with field orientation present through $\Omega_2=\pm\omega_L$3, $\Omega_2=\pm\omega_L$4, and the induced supercurrent rather than through a detailed polarization-selection-rule analysis (Mootz et al., 2023). This suggests that polarization-resolved MDCS is best regarded not as a separate spectroscopy from MDCS, but as a tensor- and symmetry-selective extension of the same multidimensional framework into increasingly diverse nonlinear-response settings.

Within that broader framework, MDCS has already been presented as a mature and broadly useful optical characterization tool for $\Omega_2=\pm\omega_L$5-conjugated polymers, capable of resolving information inaccessible to linear absorption and emission, separating homogeneous from inhomogeneous disorder, measuring coupling pathways directly, and operating in fluorescence, photocurrent, transient-absorption, and device-relevant modes (Gutiérrez-Meza et al., 2022). Polarization resolution adds the further ability to engineer which pathways, symmetries, or operator sectors remain visible in the multidimensional spectrum.