Content
Citation(s) & Footnote(s)
Book(s)September 28 fresh coverage audit — complete
Fresh sequential reading now covers all 533 supplied PDF sheets, including the entire appendix, bibliography and both indexes. Visual coverage is complete: 429 enlarged pages and six overview grids of the other 119 pages (15 subsequently enlarged). All 21 linked Excerpts, 21 Citations and nine Zettels have been audited; repaired records were read back and full supplied-source processing is verified. The audit does not independently validate all underlying primary studies. The original, attachment and existing evidence identities are preserved.
Chapter 2 needs substantive qualifications before Academy reuse. PDF 88–89 (printed 77–78) incorrectly interprets p-values as the probability that the null hypothesis is true; significance does not mean accepting the null with 5% confidence. For 20 independent triangle trials with chance success 1/3, exact upper-tail probabilities for 11, 13 and 14 correct responses are approximately 0.03764, 0.003725 and 0.000879. Those count cutoffs fit the stated significance levels, but the surrounding interpretation is incorrect. Trial independence and multiple comparisons also need consideration.
PDF 89–92 (printed 78–81) reports failure to distinguish sampled whiskies; this is not proof of sensory equivalence or that mashbill never matters. The R-index is a rank-discrimination measure, and its p-value is not the probability that a claim is wrong. Retain the specific panels, bottlings, sensory route and experimental conditions. Reconstructed Jim Beam aroma was assessed orthonasally, not shown to reproduce drinking experience.
Image checks confirm an inconsistent Gaussian expression on PDF 64 (printed 53): the displayed diffusion equation requires an exponent denominator of 4Dt and a spatial standard deviation containing sqrt(2), whereas the printed solution uses 2Dt. PDF 79 (printed 68) prints a dilution index of 8196 three times despite a powers-of-two protocol and repeats the same substituted pyrazine under reordered names with different values; primary reconciliation is pending. PDF 74's concentration axis is micrograms/L and error bars are standard deviations.
Malting figure checks confirm swapped captions for Figures 3.11–3.12 (PDF 104–105) and implausibly transposed axis labels in Figure 3.20 (PDF 109). Table 3.2 (PDF 117) prints pg/L where the neighboring table uses micrograms/L; primary reconciliation is required. Figure 5.1 (PDF 155) incorrectly says step XIV forms ethanol although its arrow ends at acetaldehyde. Fermentation tables show compound-specific temperature responses, not a universal ester optimum. Distinguish bacterial culture from distilled backset when explaining sour mash. Early distillation experiments support treating separation, reaction and copper contact as interacting processes; historical glass-still results are not universal commercial operating prescriptions.
Additional figure checks through PDF 172 identify unreliable biochemical labels: Figure 5.9 reverses the apparent serine/O-acetylserine arrangement and misnumbers a step; Figure 5.10 defines alpha-KGA inconsistently with its glutamate product; Figure 5.11 draws hydroxypropionic acid converting directly to acrolein without a reduction step. Primary reconciliation is required before reusing these pathways. Distillation theory through PDF 239 usefully separates instantaneous vapor from cumulative spirit and explicitly models recycled feints at steady state. Its normalized congener concentrations, ideal equilibrium stages and fixed cut endpoints must not become brand predictions or elapsed-time claims. Reflux may change separation and energy demand; its graph does not establish how fast a real still runs.
Maturation reading through PDF 330 supports conditional explanations of barrel preparation, entry proof, reused casks and filtration. Separate extraction from reaction and evaporative concentration; the six-liter versus 180-liter cask experiment used different durations and cannot supply a universal accelerated-aging ratio. Historical assays, class-equivalent measurements, peak areas and sensory scores are not interchangeable. PDF 315 misidentifies a 70 ng/g octenal result as nonenal in ppm; PDF 323 conflicts between a four-year prose description and a six-year figure caption. Image inspection confirms these discrepancies; primary reconciliation remains open. Doubler equations on PDF 234–236 also contain inconsistent incoming-stream labels and a missing differential factor; do not implement them without deriving the conservation balances. Model distillation energy balances do not by themselves establish the reversible minimum separation work.
Spirit-matrix reading through PDF 373 distinguishes measured nonideal behavior from proposed molecular explanations. Activity coefficients must be combined with changing congener concentration; their increase alone does not establish a stronger aroma or greater preference after dilution. Conflicting aging/matrix studies do not yield a molecular age clock or universal resting time. The PDF367 dodecanal table contradicts an intermediate-response claim in the prose. Frozen-sample calorimetry does not establish an optimal room-temperature tasting proof. Volume, compressibility and viscosity studies offer differing structural interpretations; negative Gibbs energy of mixing alone is insufficient to demonstrate stability against separation.
Gauging (PDF 374–426) explains why shape, reference temperature, contraction and non-ethanol solutes limit simple proof estimates. Historical bead tests are affected by congeners and do not establish strength or purity. Several numerical and historical inconsistencies need primary reconciliation before reuse. Appendix A (fully read and visually examined, PDF 427–473) preserves olfactory versus taste and recognition versus detection thresholds, solvent composition, inequalities and primary references. Its heterogeneous values are not interchangeable universal thresholds; the isoamyl-alcohol headspace hypothesis is explicitly only qualitative. Preserve solvent, detection/recognition/taste/astringency endpoint, stereochemistry, inequalities and cited experiment separately. PDF466 labels both whisky-lactone 300 and 7000 ppb observations as detection; several chemical synonyms and lactone stereochemical assignments need primary reconciliation. Copper is 2.4–3.8 mg/L in PDF178 but 2.4–3.8 under the appendix ppb heading in PDF473, an internal factor-of-1000 inconsistency. Thresholds are not exposure limits. The chemical index repeats questionable names and must not be imported as an authoritative synonym dictionary.
Proposed Academy uses: compare aroma-screening methods with recombination/omission validation; train with explicitly defined references; use blinded category exercises to explore overlap; explain environmental expectations without treating color as physical-age evidence. Detailed page-level notes and unresolved checks are saved in the local audit. No public lesson changes have been made.
This is the canonical full-book Literature Note for Gregory H. Miller's Whisky Science: A Condensed Distillation. It supersedes the earlier pilot-only synthesis without creating a duplicate.
Author's argument
Whisky is a coupled system. Historical institutions define product identity; malting, mashing, fermentation, and distillation create and select precursors; maturation adds, removes, and transforms compounds at different rates; the ethanol-water matrix changes their activity; and human perception turns that moving chemical system into flavor. Miller argues that useful explanations must follow mechanisms and conditions rather than rely on category labels or single measurements.
Chapter synthesis
History and identity
American whiskey emerged through relationships among distillers, rectifiers, merchants, tax law, bonded storage, consumer trust, and standards of identity. Prolonged storage became economically normal before it was framed solely as a quality tradition, and the charred-barrel inventor stories are less secure than popular retellings suggest.
Flavor and sensory method
Most apparent taste is retronasal aroma, and measurement is shaped by matrix, expectation, protocol, and the assessor. Thresholds and odor-activity values help prioritize candidates but do not predict complex mixtures. A trained-panel study in the reviewed literature did not find an attribute that reliably separated its bourbon and rye samples.
Malting and mashing
Malting is controlled germination and kilning; peat-smoke phenols are absorbed by moist malt during kilning rather than supplied by peaty water. Barley dominates the evidence base, so translation to corn and rye requires care. Mashing sets yield and flavor through gelatinization, enzyme activity, lipids, phenols, minerals, temperature, and on-grain practice.
Fermentation
Yeast and bacteria jointly influence acids, esters, higher alcohols, sulfur compounds, lactones, and defects. Sour mash is best understood as pH control, continuity, and managed microbial ecology rather than a whiskey style.
Distillation
Vapor-liquid equilibrium matters, but ideal models omit reactions, residence time, holdup, foam and mist carryover, and changing copper chemistry. Direct fire and steam heat provide different reaction pathways. The causal unit is the operating system—not simply “pot” or “column.”
Maturation
Wood extraction, lignin ethanolysis, oxidation, esterification, volatilization, and sorption advance at different rates. Oak anatomy, cutting, seasoning, toast, char, barrel size, entry proof, oxygen, warehouse, and spirit all matter. Small barrels accelerate variables unequally; higher entry proof generally reduced several measured maturation markers in the cited American studies; a special requirement for post-blend marriage in wood remains unproven.
Spirit matrix, gauging, and thresholds
Ethanol and water form a nonideal matrix that changes congener activity and headspace partitioning. Dilution can selectively increase the aroma availability of long-chain esters. Hydrometers and historical proof systems infer strength through physical properties and administrative conventions. Threshold values are conditional on solvent, temperature, protocol, definition, and assessor population.
Evidence architecture — September 28 audit complete
- Source: No access
- Coverage: 533 of 533 PDF sheets reviewed textually and visually.
- Evidence records: 21 verified Excerpts with dual PDF/printed locators.
- Citations: 21 Chicago Notes and Bibliography records marked Ready.
- Durable notes: nine related Zettels, including six created from the full-book review.
- Claims: new source-qualified claims were created only when no equivalent record existed; older claims were strengthened through reciprocal relations.
Cross-book connections and proposed Academy applications
Miller and Jackson both show why threshold values must retain solvent, sensory task, units and provenance. Their tables also contain enough inconsistencies to make per-value checks essential; proposed threshold-reference cards should separate observations rather than average them. Jackson provenance: 
Read alongside Whisky Classified, Miller supplies process mechanisms for sensory patterns, but neither descriptive clustering nor a null category comparison proves a unique production cause. A proposed Academy exercise would compare blinded descriptors first, then disclose production information and discuss alternative explanations. Classified provenance:
A second proposed exercise compares the same whiskey at controlled proofs and equal serving temperatures, recording dilution, headspace time and assessors' observations. This tests the sample experience; it does not prove a molecular restructuring theory. A third uses small-cask and entry-proof diagrams to distinguish extraction, reaction and evaporation without assigning equivalent years or promising superior flavor. These remain proposed uses; no public lessons were changed.
Researcher synthesis
The website should explain American whiskey through causal chains:
- Institution and history shape what counts as whiskey and how it reaches market.
- Materials and organisms create a field of possible flavors.
- Equipment and operation select and transform that field.
- Wood, time, proof, and environment move several maturation processes at different rates.
- Matrix and perception determine which compounds become sensorially active.
This structure avoids deterministic shortcuts while remaining accessible to a general reader.
Reliability boundary
The source is valuable because it states uncertainty plainly. Proprietary work, thin replication, small samples, incomplete characterization, and cross-category evidence prevent universal numerical rules. American-specific website claims should retain their conditions and, where consequential, be corroborated with primary or independent evidence.
