Hermeneutic Calvinball versus modest digital humanities in philosophical interpretation

The paper frames a “Calvinball problem” in philosophy and especially in the history of philosophy: the lack of stable, commonly accepted interpretive rules leads to incompatible readings and weak criteria for adjudication. After noting broader pressures on the humanities and disputes over philosophical progress, the author proposes that digital humanities can supply modest, defeasible defaults that constrain interpretation. Specifically, quantitative text analysis can help determine which concepts are central to a philosopher, how their salience changes over time, and which conceptual linkages are supported by the corpus. The study applies this to Nietzsche, contrasting primary texts with secondary literature on the emotions of resentment and shame.

The paper surveys interpretive practices ranging from strict textualism to speculative readings influenced by Heidegger and Derrida, noting how both what a philosopher writes and what they did not write have been used as evidence. It critiques influential Anglophone interpretations that hinge on single passages (e.g., Swanton on “mature egoism” from HH 95; Clark and Dudrick on “will to value” from BGE 2) and argues that translator choices (e.g., italicizing ‘Ressentiment’ as if foreign) have amplified the perceived importance of resentment. Quantitative contrasts between primary and secondary literatures are highlighted: Nietzsche mentions resentment/ressentiment relatively rarely (21 passages) versus shame (159 passages across all published/authorized works), while the secondary literature disproportionately focuses on ressentiment (e.g., 149 PhilPapers hits for “Nietzsche and ressentiment” vs 14 for “Nietzsche and shame” since 1950). The paper situates these issues within broader concerns about the humanities’ status and the need for replicable methods.

The study combines quantitative digital humanities techniques with close reading:

• Corpus: Nietzsche’s published and authorized works (excluding Nachlass and letters), sourced from Nietzsche Source.
• Baseline tallies: Stem searches for resentment/ressentiment and shame (scham*, schmach*, schand*) to estimate relative concept importance.
• Periodization and stylistic overview: Hierarchical clustering using Euclidean distance on normalized token frequencies to reveal clusters (free spirit works; mature works; late works).
• Lexical dispersion: Plotting the distribution of shame-related stems across works to visualize usage over time and within texts.
• Relative frequency: A composite dictionary of shame terms to compute per-book frequency proportions across the corpus.
• Conceptual associations: Paragraph-level co-occurrence analysis to construct a conceptual collocation egonet centered on shame, identifying associated concepts (e.g., conscience, contempt, laughter, solitude, virtue).
• Close reading: Detailed analysis of the highlighted passages to interpret the functions of shame and assess how collocated concepts figure in Nietzsche’s moral psychology. All data and code (RMarkdown) are shared via OSF for reproducibility.

• Primary vs secondary mismatch: Nietzsche discusses shame in 159 passages across all published/authorized works but resentment/ressentiment in 21 passages (GM, A, TI, EH). By contrast, secondary literature emphasizes ressentiment (e.g., PhilPapers: 149 “Nietzsche and ressentiment,” 70 “Nietzsche and resentment,” 14 “Nietzsche and shame” since 1950).
• Translator effect: Systematic transliteration/italicization of ‘Ressentiment’ in English editions likely inflated its perceived importance.
• Periodization: Hierarchical clustering shows stylistic groupings (free spirit works HH, D, GS; mature BGE, GM; late EH, TI), providing context for tracking concept trends.
• Persistence of shame: Relative frequency and dispersion analyses show shame is a cross-period concern (notably in RWB, D, Z, BGE, GS, A).
• Conceptual collocations with shame: conscience, contempt, laughter, solitude, and virtue occur frequently in shame-related paragraphs, guiding interpretive focus.
• Four functions of shame in Nietzsche:
  1. Social regulation among (near-)equals: shame/disgrace helps maintain equilibria by deterring norm violations (e.g., HH WS 22; GM 2.5 on creditor–debtor relations).
  2. Pathos of distance (virtue): a refined, prospective sense of shame that tracks rank and prevents shaming oneself or others; includes the “shame of a god” (BGE 40), critiques of scholars’ lack of shame (BGE 263; GS 358; EH Wise.4; Z 2.Pity), and maxims on sparing shame (GS 273–5).
  3. Vicious shame: pathological when directed at fixed aspects of human nature or the self (e.g., ascetic self-shaming, D 109; Z 1.Warriors; Z 4.Ugliest; GM 2.7 on species-level shame; BGE 195; A on Pauline shaming of wisdom/power).
  4. Counter-shame: Nietzsche shames promoters and victims of vicious shame, sometimes to uplift those who should know better (RWB 11; GM 1.11; Z 3.Apostates2; A 26, 38, 62).
• Overall: Shame is complex, inescapable, and function-relative; apt shame can be self-effacing and regulative, while inapt shame is degenerative.

The findings substantiate the paper’s proposal that modest digital humanities can curb hermeneutic arbitrariness by offering empirical defaults about concept importance, periodization, and conceptual linkages. In Nietzsche’s case, quantitative evidence corrects the secondary literature’s overemphasis on ressentiment and foregrounds shame as a sustained thematic concern across periods. Collocation results focus interpretive attention on conscience, contempt, laughter, solitude, and virtue, which close reading confirms as central to Nietzsche’s treatment of shame. The four-function account explains how shame supports social equilibria, matures into the pathos of distance, devolves into pathology when misdirected at fixed aspects of nature/self, and is combated by Nietzsche’s counter-shame. Methodologically, the study illustrates how combining scalable text analysis with targeted close reading can expose cherry-picking, align interpretations with corpus-wide patterns, and enhance replicability in the history of philosophy.

The paper contributes a case for replacing “hermeneutic Calvinball” with modest, replicable digital humanities practices. Empirically grounded analysis shows shame, not ressentiment, is a major, persistent concern in Nietzsche’s published/authorized works and that it is conceptually intertwined with conscience, contempt, laughter, solitude, and virtue. Interpretively, Nietzsche treats shame as function-relative: beneficial for social regulation and the cultivation of the pathos of distance, harmful when aimed at immutable aspects of human nature or the self, and appropriately resisted via counter-shame. Future work should scale these methods to other philosophers, concepts, and corpora; integrate richer multilingual and variant-text analyses; and develop standardized protocols for collocation, periodization, and replication to strengthen methodological norms in philosophical interpretation.

• Quantitative counts and co-occurrences provide defaults, not definitive meanings; they require corroboration via close reading.
• Search-based measures (stems scham*, schmach*, schand*) may miss nuance or context and depend on textual normalization and edition choices; Nachlass and letters are excluded.
• Collocation indicates association, not causation or directionality; interpretive significance must be argued textually.
• Translator/editorial practices affect Anglophone reception and frequency patterns (e.g., treatment of ‘Ressentiment’), potentially biasing secondary literature comparisons.
• Periodization via hierarchical clustering is sensitive to preprocessing and feature choices; stylistic clusters may not perfectly track conceptual development.
• The approach is demonstrated on a single case study (Nietzsche) and focuses on specific emotions; generalizability requires further applications.