A sensory threshold belongs to a compound-plus-matrix-plus-method-plus-person context; it is not a universal constant of the molecule.
Threshold compilations remain useful for screening, but only when their conditions and uncertainty travel with the value.
Threshold values change across water, beer, wine, and spirits and across detection, recognition, olfactory, and taste protocols. Retrieval must preserve the solvent and method with the number.
Created after a full 533-sheet textual and visual review of Whisky Science (2019); source-specific limits retained. September 28, 2026: Fresh Miller audit: all Appendix A tables were read and visually inspected. Store stereochemistry, solvent, endpoint, inequalities and source alongside each value. Copper has a factor-of-1000 internal unit discrepancy (PDF178 versus473); whisky-lactone detection labels and several chemical aliases require reconciliation. Together with Jackson's table issues this supports checking each observation, not blindly importing a threshold table.
One durable idea synthesized from the complete review of Whisky Science: A Condensed Distillation.
Atomic idea
A threshold is conditional on molecule, matrix, method, temperature, and assessor—not an intrinsic universal number.
Reasoning
The appendix records large differences across water, beer, wine, and spirit matrices and across detection, recognition, olfactory, and taste procedures. A usable threshold record must retain those conditions, because headspace partitioning and human perception both change with context.
Provenance
September 28 source audit
Fresh Miller audit: all Appendix A tables were read and visually inspected. Store stereochemistry, solvent, endpoint, inequalities and source alongside each value. Copper has a factor-of-1000 internal unit discrepancy (PDF178 versus473); whisky-lactone detection labels and several chemical aliases require reconciliation. Together with Jackson's table issues this supports checking each observation, not blindly importing a threshold table.
Counterpoint
Thresholds still help prioritize candidates, but they do not predict mixture behavior by themselves and should not be detached from their experimental conditions.
Maturity checklist
Jackson 2017 — September 28 source audit
Jackson's Table 3.2 and surrounding discussion (supplied PDF117–123) reinforce recording matrix and method with every threshold. The fresh review also found implausible printed units and duplicate chemical aliases; a table's authority cannot replace verification of the particular value. Proposed Academy references should retain source, matrix, unit, detection/recognition task and uncertainty. No universal whiskey threshold is inferred.
Provenance: