{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"negative_space_design__astronomy_astrophysics","trajectory_id":"R","attempt_index":0,"archetype_slug":"negative_space_design","domain_slug":"astronomy_astrophysics","decision":"CANDIDATE","problem_id":"spectral_line_baseline_without_clean_reference_channels","causal_lever_id":"protected_line_free_spectral_windows","proposal":{"problem":"Spectral-line observations can leave too few demonstrably line-free channels around a target because broad line wings, neighboring transitions, radio-frequency interference, and band-edge artifacts occupy the available band. Baseline or continuum subtraction then absorbs real emission or preserves instrumental structure, biasing line fluxes, widths, and detections. [INFERENCE]","actors_substrate":["observing-program principal investigators","telescope and spectrometer operators","spectral-line data analysts","archive and catalog users","channelized radio, millimeter, or submillimeter spectra"],"observable_state":"A target-line band has no predeclared, independently checked reference windows on both sides of the expected signal; baseline estimates change materially when plausible channel masks or polynomial orders change.","consequence":"Instrumental or continuum structure becomes underconstrained and confounded with astrophysical emission, producing unstable measurements or false features. [HYPOTHESIS]","affected_objective":"Estimate faint spectral-line presence, integrated flux, width, and profile without mistaking baseline structure for emission or subtracting broad real signal.","structural_mapping":[{"archetype_element":"Deliberate absence","domain_realization":"Spectral intervals expected to contain neither the target line nor known neighboring lines are allocated and protected as reference windows. [HYPOTHESIS]","claim_kind":"HYPOTHESIS"},{"archetype_element":"Crowding that obscures positive form","domain_realization":"Target emission, other transitions, RFI, atmospheric features, and edge artifacts compete for the same finite channel band. [INFERENCE]","claim_kind":"INFERENCE"},{"archetype_element":"Positive form clarified by surrounding emptiness","domain_realization":"Clean reference windows constrain the baseline against which the target-line profile is measured. [HYPOTHESIS]","claim_kind":"HYPOTHESIS"},{"archetype_element":"Bounded and recoverable omission","domain_realization":"Contaminated channels are masked with provenance while raw spectra remain intact and masks can be revised.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Meaningful empty state","domain_realization":"A channel with no detected emission is distinguished from missing, flagged, insensitive, or calibration-failed data.","claim_kind":"INFERENCE"}],"component_map":[{"component":"Omission Candidate","status":"adapted","domain_realization":"Identify RFI-contaminated, artifact-dominated, or known interloper channels for reversible exclusion from baseline fitting."},{"component":"Protected Empty Space","status":"adapted","domain_realization":"Reserve line-free reference windows, preferably bracketing the expected target profile, and prohibit routine signal-model expansion into them."},{"component":"Positive Form Relationship","status":"adapted","domain_realization":"Use the reserved windows specifically to define the local baseline supporting measurement of the intervening target line."},{"component":"Attention Competition Map","status":"adapted","domain_realization":"Map expected target extent, neighboring transitions, RFI occupancy, atmospheric structure, and band-edge instability before fixing windows."},{"component":"Absence Boundary","status":"adapted","domain_realization":"Record exact frequency or velocity limits and minimum usable-channel requirements for each reference window."},{"component":"Clarity or Effect Test","status":"adapted","domain_realization":"Test whether held-out residuals and recovered injected-line parameters become more stable than under the baseline."},{"component":"Rest and Pacing Zone","status":"adapted","domain_realization":"Distribute reference intervals around dense spectral regions so baseline behavior is sampled before and after the target rather than only at one edge."},{"component":"Meaning-of-Absence Check","status":"adapted","domain_realization":"Verify that apparent emptiness reflects sensitivity-compatible non-detection, not flags, data loss, calibration failure, or unmodeled broad wings."},{"component":"Reintroduction Trigger","status":"adapted","domain_realization":"Reclassify or reacquire a reference window if residuals, ancillary data, or expanded velocity expectations show emission within it."},{"component":"Accessibility and Recoverability Guardrail","status":"adapted","domain_realization":"Preserve raw channels, masks, reasons, calibration versions, and an unmasked view so exclusions are auditable and reversible."},{"component":"Context Preservation Frame","status":"adapted","domain_realization":"Retain frequency coordinates, spectral resolution, beam, sensitivity, calibration uncertainty, target redshift or velocity, and mask metadata."}],"mechanism_dispositions":[{"slug":"architectural_void","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Treat bracketing line-free windows as deliberately sized spectral voids whose boundaries support the target-line measurement.","counterfactual_removal":"Without protected reference windows, baseline and broad emission remain confounded; the proposed causal chain loses its identifying contrast."},{"slug":"blank_or_rest_frame","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"A presentation pause does not improve spectral identifiability; temporal pacing is not the constraint.","counterfactual_removal":"No material change."},{"slug":"editorial_cut","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Reversibly mask contaminated channels while testing that retained channels preserve the context needed for baseline estimation.","counterfactual_removal":"Windows could still work, but contamination handling would be less reproducible and potentially destructive."},{"slug":"empty_state_design","disposition":"selected_supporting","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Label no emission, missing data, flagged data, and calibration failure as distinct states.","counterfactual_removal":"Apparent clean channels could be accepted because data are absent or unusable, invalidating the reference."},{"slug":"facilitation_silence","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"The mechanism depends on social withholding and participant response, absent from channelized measurement.","counterfactual_removal":"No material change."},{"slug":"focus_mode_or_control_hiding","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Hiding channels aids display focus but does not create independent baseline information; masks are already handled by editorial_cut.","counterfactual_removal":"No material change."},{"slug":"margin_and_gutter_system","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Encode reusable minimum widths and separations for target exclusion zones and reference windows, allowing instrument-specific scaling.","counterfactual_removal":"The core effect remains possible, but window protection becomes ad hoc and vulnerable to erosion."},{"slug":"negative_space_logo","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"Figure-ground symbolic double reading has no role in estimating a physical spectral baseline.","counterfactual_removal":"No material change."},{"slug":"pause_in_speech","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"Live rhetorical pacing does not map to the measurement substrate.","counterfactual_removal":"No material change."},{"slug":"sparse_layout","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Reducing displayed spectral elements might improve inspection but cannot substitute for clean observed reference channels.","counterfactual_removal":"No material change."},{"slug":"whitespace","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Perceptual spacing is only a visualization aid; architectural_void already captures the load-bearing spectral allocation.","counterfactual_removal":"No material change."}],"causal_chain":["Predeclare reference-window boundaries using expected line extent and a contamination map.","Observe enough bandwidth and sensitivity for usable reference channels on both sides of the target where feasible.","Diagnose each apparently empty channel and reversibly exclude contaminated or invalid samples.","Fit the baseline from protected windows without allowing the target model to consume them by default.","Compare held-out residuals and injected-signal recovery across window definitions.","If stable, subtract the constrained baseline and estimate target-line properties with mask and window provenance attached."],"baseline":"Use the available band primarily for target coverage, choose apparently line-free channels after inspection, and fit a polynomial or continuum model whose order and mask are adjusted until residuals appear acceptable.","nearest_rival":"Fit a flexible joint model of baseline, contamination, and astrophysical line across the full band without reserving protected reference windows.","authority_safety":{"affected_parties":["observers sharing finite telescope time or bandwidth","analysts interpreting the spectra","downstream catalog and archive users","scientists studying broad or unexpected spectral features"],"decision_authority":"The observing-program principal investigator and designated calibration or instrument scientist may define pilot windows, subject to observatory configuration and data-quality rules.","authorized_first_step":"On one existing spectral dataset, pre-register two plausible bracketing reference-window schemes, preserve all raw channels, inject a bounded grid of synthetic lines, and compare recovery bias and held-out residuals against the ordinary baseline and nearest rival.","excluded_actions":["irreversibly delete excluded channels","label a window line-free solely because no feature is visually obvious","hide masks or calibration failures from downstream users","sacrifice required safety, calibration, or shared-observer constraints","generalize from the pilot to other instruments or line regimes without retesting"],"halt_rollback":"Halt if either protected window shows statistically structured residuals, plausible broad-line occupancy, excessive data loss, or worse injected-line recovery than both comparators. Revert to the untouched spectra and recorded original masks; expand, relocate, or abandon the windows."}},"negative_tests":{"strongest_counterevidence":"A sufficiently accurate joint physical-instrument model may recover unbiased line parameters without dedicated line-free windows, while reserved bandwidth can reduce velocity coverage or sensitivity to unexpected features. [HYPOTHESIS]","analogy_break":"Astronomers cannot manufacture an emission-free part of the sky: a planned spectral void may contain unknown lines, broad wings, or RFI. Unlike graphic whitespace, its emptiness is an empirical proposition that must be repeatedly tested.","failure_condition":"The method fails when no reference interval is demonstrably cleaner than the target region, when the baseline is nonlocal or discontinuous across the windows, or when protecting windows excludes scientifically important broad or unexpected emission.","problem_falsifier":"Across representative spectra, line estimates remain stable under plausible masks and baseline orders, clean reference channels are already reliably available, and baseline uncertainty is negligible relative to other errors; then lack of protected spectral space is not the problem.","intervention_falsifier":"Given spectra exhibiting the diagnosed instability, predeclared protected windows do not improve held-out residuals or injected-line recovery relative to both the ordinary baseline and joint-model rival, or they increase total measurement error through lost coverage.","risks":["Weak or broad real emission may be misclassified as baseline and subtracted.","Known-line maps may create confirmation bias against unexpected discoveries.","Extra bandwidth or observing time may displace other science objectives.","Rigid standard windows may fail across redshift, instrument, or source class.","Clean-looking channels may actually be insensitive, missing, or calibration-failed."]},"null_rationale":null,"classification":{"candidate_kind":"MECHANISM_ADAPTATION","prior_art_status":"UNSEARCHED","evidence_maturity":"HYPOTHESIS"},"revision_change_log":{"revision_kind":"ORIGINAL","prior_problem_id":null,"prior_causal_lever_id":null,"problem_changed":false,"causal_lever_changed":false,"conceptual_changes":[],"operational_changes":[],"repairs_addressed":[]},"confidence":0.76,"generator_notes":"Closed-book structural transfer. The candidate is limited to spectral regimes where reference-window emptiness is testable; no novelty or prevalence claim is made."}