Case Selection

Choose a case

American Revolution

Revolutionary founding, sacred liberty, collective exception, and war.

Napoleon / Napoleonic Wars

Imperial glory, army–nation fusion, and translation-sensitive sacrifice.

French Revolution / Jacobin Republic

Draft case for virtue, terror, public safety, and sacred republican peoplehood in French source text.

British World War I / Lloyd George

Draft limiting/control case for democratic wartime sacrifice, endurance, liberty, and anti-autocracy rhetoric in English OCR source text.

Lincoln / American Civil War

Union, covenant, battlefield sacrifice, providence, and reconciliation.

Hitler / Nazi Germany

Purification fantasy, enemy construction, and exterminatory violence.

Why these cases

V1 uses leader-centered corpora to compare how sacred objects, enemies, obligation, and sacrifice are organized in political language. The cases are not treated as identical historical events. They are selected because each can test some part of the Law of Sacrifice across war, revolution, empire, civil war, and genocide.

The initial pilot cases for the primary research question are defined here rather than in the research-question document so that the question can remain stable as the workbench expands.

Selection criteria

Cases should satisfy most of these criteria:

enough public-domain or open-access material for a balanced core;
clear connection between political violence and sacred collective objects;
enough phase coverage to test diachronic development;
enough register variation to test rhetorical setting;
enough historical difference to make cross-case comparison meaningful;
source provenance clear enough for citation and audit.

Pre-v1 expansion rubric

Status: expansion criteria

These criteria govern the bounded pre-v1 corpus expansion window tracked in GitHub milestone 7 and issue #174. They should be applied before acquiring new source text. Candidate ranking and final selection happen in issue #176.

The expansion window may add new cases or selected documents only when they improve the evidentiary design enough to justify added rights review, annotation, reliability sampling, and publication-status work. Expansion should test the framework, not simply add more examples that resemble the existing cases.

The ranked candidate matrix for this window is recorded in Pre-v1 Expansion Candidate Matrix. Rights and provenance decisions for those candidates are recorded in Pre-v1 Expansion Rights and Provenance Review.

Case-level rationales

Each proposed new case should have at least one primary rationale and may have secondary rationales:

Rationale	Use when the case would…
Confirmatory	Test whether a strongly sacrificial or purification-centered corpus produces the expected Koenigsbergian pattern.
Limiting	Test a case where political violence is present but sacrificial, body-politic, or enemy-death language may be weak, displaced, or absent.
Control	Test a case with mobilization or crisis rhetoric that should not be expected to support the Law strongly.
Comparative	Add historical, institutional, religious, ideological, geographic, language, genre, or regime-type contrast not covered by the current four cases.
Source-availability	Add a case because rights-cleared, well-provenanced sources make it unusually suitable for reproducible method testing.

A source-availability rationale is not sufficient by itself. It can break ties between otherwise valuable candidates, but it should not make a weak comparative case enter the corpus.

New-case threshold

A new case should be admitted only when it satisfies all mandatory criteria and scores well on most value criteria.

Mandatory criteria:

Stable source base with citable provenance.
Rights status compatible with the intended storage, publication, or metadata-only policy.
At least one analyzable leader-centered or movement-centered corpus layer.
Clear relevance to one or more research questions.
Feasible annotation burden before the expanded-corpus freeze.
Explicit language and translation policy.
Expected contribution that is not already supplied by Lincoln, American Revolution, Napoleon, or Hitler.

Value criteria:

Improves variation across war, revolution, empire, civil war, genocide, authoritarianism, democracy, colonialism, religion, or anti-colonial struggle.
Adds a strong positive, limiting, or control test rather than another near-duplicate positive case.
Adds a non-English or non-Western corpus only when translation and source controls are strong enough for annotation.
Adds genre or register contrast, such as speeches, orders, manifestos, sermons, proclamations, diaries, or bureaucratic implementation language.
Clarifies the war/genocide distinction, the enemy-as-bringer-of-death mechanism, sacred-object formation, or sacrifice without enemy destruction.
Supports a balanced core that can be compared without hiding major source asymmetries.

Document-level threshold

Additional documents for existing cases should be admitted only when they fill a defined gap. A document is necessary when it does at least one of the following:

Covers a missing phase, turning point, genre, register, audience, or rhetorical occasion.
Tests a known absence or weak point in the current case analysis.
Repairs a source-quality, provenance, translation, OCR, or coverage problem.
Adds negative evidence needed to prevent overclaiming.
Supports a balanced-core comparison that would otherwise be distorted.

A document should be excluded or deferred when it is merely illustrative, duplicative, background-only, too rights-risky for the intended use, too fragmentary for reliable annotation, or likely to expand one case without a clear comparative reason.

Ranking fields

Candidate-selection work should record these fields for every proposed case or document:

Field	Purpose
Candidate ID	Stable short identifier for discussion and later manifest work.
Proposed layer	New case, existing-case balanced core, existing-case extended corpus, or metadata-only context.
Primary rationale	Confirmatory, limiting, control, comparative, or source-availability.
Research-question link	Which primary or secondary research question the candidate tests.
Source base	Candidate editions, repositories, transcripts, scans, or local-only sources.
Rights risk	Low, medium, high, or blocked, with storage implications.
Language policy	Source language, translation aid, gloss policy, and contested terms.
Genre/register value	What rhetorical setting the candidate adds.
Annotation cost	Low, medium, high, or prohibitive, considering lexical volume and language.
Reliability impact	Whether human or model reliability sampling must expand.
Comparative gain	What the project learns that the current four-case corpus cannot test.
Exclusion risk	The strongest reason to reject or defer the candidate.
Recommendation	Admit, defer, metadata-only, reject, or needs rights review.

Expansion stop rules

The expansion window should stop before acquisition or annotation continues when any of these conditions hold:

The candidate set no longer changes the comparative design.
Rights or translation risk would force claims beyond what the project can responsibly audit.
Added annotation volume would delay reliability and publication promotion without a corresponding methodological gain.
A proposed source would require unbounded redownload, source harvesting, or a broad reset of already verified corpora.
The expansion would make balanced-core comparison less transparent rather than more informative.

Comparative design

Case	ID	V1 corpus model	Rationale
American Revolution	`am-rev`	Washington-anchored collective exception	Tests sacrifice, founding, sacred liberty, and collective exception in revolutionary war language.
Napoleon / Napoleonic Wars	`napoleon`	Napoleon-only individual leader corpus	Tests imperial glory, leader-body extension, army/nation fusion, and translation-sensitive sacrifice language.
French Revolution / Jacobin Republic	`fr-rev`	Robespierre-centered Jacobin public corpus	Draft expansion case testing republican virtue, terror, public safety, sacred peoplehood, and source-with-glosses French annotation.
British World War I / Lloyd George	`wwi-britain`	Lloyd George-centered democratic wartime leadership corpus	Draft expansion case testing sacrifice, endurance, democratic wartime mobilization, liberty, and anti-autocracy rhetoric without genocidal enemy destruction.
Lincoln / American Civil War	`lincoln`	newly selected Lincoln corpus	Tests body-politic healing, covenant, experiment, providence, war sacrifice, and systematic absence.
Hitler / Nazi Germany	`hitler`	Hitler-centered corpus with limited Goebbels/Goering support if needed	Tests the war/genocide boundary, purification fantasy, enemy as bringer of death, and exterminatory body-politic logic.
Cross-case synthesis	`x-case`	synthetic synthesis workspace	Not a historical case; receives case-level outputs for comparison, validation, and synthesis.

Limits

Case selection does not imply moral equivalence, identical causation, or a single psychological mechanism. Cross-case comparison should identify both shared symbolic operations and historically specific differences.

The case set remains provisional until source discovery confirms that each case can support a balanced core and the necessary validation checks.