VOLKSWAGEN DIESELGATE DEMO CORPUS DOSSIER
Status: RESEARCH / PRE-CAPTURE ONLY
Status: RESEARCH / PRE-CAPTURE ONLY
Subject: Volkswagen "Dieselgate" — U.S. defeat-device conduct, real-world emissions discovery, regulatory findings, criminal admissions, consumer/environmental settlements, vehicle-scope distinctions, management-knowledge claims, and modeled health impacts
Research date: 2026-08-09
Scope: source inventory, authority/posture, contradiction/qualification map, provisional five-claim matrix, capture hazards, reverse-source candidates
Explicit non-claims: This dossier creates no CaptureReceipt, SourceEdition, SRS receipt, custody, admission, final standing, article prose, or canonical Counterpedia identity.
A. SUBJECT / DISAMBIGUATION
Common name
Volkswagen Dieselgate / Volkswagen emissions scandal
Exact scope
Volkswagen Group's use in U.S.-market diesel vehicles of software/calibrations that caused emissions controls to perform differently during regulatory testing than during normal driving; the independent real-world testing that surfaced the discrepancy; EPA/CARB Notices of Violation and recall/remedy proceedings; Volkswagen AG's U.S. criminal guilty plea; individual criminal proceedings; U.S. Clean Air Act and consumer settlements; and scientific efforts to estimate excess NOx and health consequences.
The U.S. corpus is the first governed demo scope because it contains unusually strong adjudicative and regulatory records. Global EA189 vehicle counts and Volkswagen's current first-party legal position are included only where needed to demonstrate scope/knowledge distinctions.
This dossier is not:
a full history of diesel emissions regulation worldwide;
a claim that U.S. law and European law define defeat devices identically;
a final adjudication of every Volkswagen executive's knowledge;
an assertion that every affected vehicle emitted the same multiple of the standard;
a claim that modelled premature deaths are individually identifiable victims.
Important identifiers / terms
Naming / semantic traps
"Volkswagen cheated emissions tests" is strongly settled in the U.S. record, but "every diesel Volkswagen cheated" is false at that scope. Exact engine, model-year and test-group boundaries matter.
WVU's 2014 study tested three vehicles; it did not itself prove a fleetwide intentional defeat-device scheme. Regulators performed the later investigation that established the prohibited software conduct.
Approximately 590,000 affected U.S. vehicles ≠ roughly 11 million EA189 vehicles worldwide. Different jurisdiction/engine/vehicle populations.
Consumer compensation dollars ≠ environmental mitigation dollars ≠ criminal penalty ≠ civil penalty.
A buyback/remedy settlement is not a measure of public-health damage.
A peer-reviewed estimate of premature mortality is model-dependent. It is not a registry of identified persons shown to have died because of a specific VW vehicle.
NOx emissions are not the same variable as NO2 exposure, ozone, PM2.5, hospital admissions, or deaths. Health modelling contains transformation steps.
An approved emissions modification changes the safety/compliance-relevant configuration of a subject vehicle. "Same model" does not mean "same emissions configuration" before and after remedy.
Temporal bounds
Discovery/testing: 2013–2015
U.S. regulatory/enforcement peak: 2015–2017
Remedy/consumer program: 2016–2020+
Current corporate/legal reporting: through 2026
Why fertile for Counterpedia
Dieselgate is the ideal settled-core / disputed-scope demo.
The U.S. core is unusually firm:
real-world emissions discrepancies were measured;
regulators established prohibited defeat-device software;
Volkswagen AG pleaded guilty to felony counts arising from a long-running U.S. emissions-cheating scheme;
individual employees also pleaded guilty;
consumer and environmental remedies were imposed.
But a large halo of commonly repeated claims needs qualification:
which vehicles;
which engine families;
how many vehicles;
10×, 35×, 40×, or 9× relative to what;
who knew and when;
whether top management knew from the beginning;
how many premature deaths were caused;
whether a repaired vehicle remains equivalent to its pre-remedy edition;
whether U.S. illegality can be silently globalized to every jurisdiction.
This is precisely the kind of topic where a settled scandal can still be badly described.
B. SOURCE INVENTORY
VW-S01 — EPA September 18, 2015 Notice of Violation, 2.0L
Exact title: Notice of Violation to Volkswagen AG, Audi AG, and Volkswagen Group of America, Inc.
Issuer: U.S. Environmental Protection Agency
Date: September 18, 2015
Exact PDF locator:
https://www.epa.gov/sites/production/files/2015-10/documents/vw-nov-caa-09-18-15.pdf
Current EPA discovery page:
https://www.epa.gov/enforcement/notices-violation
Authority / posture: Clean Air Act enforcement Notice of Violation; regulator allegation/finding at enforcement stage, not criminal conviction.
May support
precise 2.0L model-year/test-group scope;
EPA's defeat-device theory;
certification-test vs normal-operation behavior;
regulatory requirements and violation notice.
Cannot establish alone
criminal guilt;
knowledge of every executive;
individual consumer damages.
Captureability: HIGH
Priority: P0
VW-S02 — EPA/CARB September 18, 2015 public violation announcement
Exact title: EPA, California Notify Volkswagen of Clean Air Act Violations
Issuer: EPA / CARB
EPA locator:
https://www.epa.gov/archive/epa/newsreleases/epa-california-notify-volkswagen-clean-air-act-violations-carmaker-allegedly-used.html
CARB locator:
https://ww2.arb.ca.gov/news/epa-california-notify-volkswagen-clean-air-act-violations
Date: September 18, 2015
Authority / posture: regulator enforcement announcement; allegations at that date.
May support
public discovery chronology;
2009–2015 four-cylinder diesel scope;
regulator characterization of software that circumvents emissions standards.
Captureability: HIGH
Priority: P1
VW-S03 — CARB September 18, 2015 In-Use Compliance Letter / archive
Issuer: California Air Resources Board
Archive surface:
https://ww2.arb.ca.gov/volkswagen-information-document-archive
Authority / posture: California in-use compliance/regulatory record.
May support
CARB's testing/investigation chronology;
California-specific certification/recall posture;
exact regulatory correspondence preceding later recall rejection/remedies.
Captureability: HIGH for archive; child letter should be captured separately.
Priority: P0
VW-S04 — WVU CAFEE / ICCT 2014 real-world testing report
Exact title: In-Use Emissions Testing of Light-Duty Diesel Vehicles in the United States
Authors: Gregory J. Thompson, Daniel K. Carder, Marc C. Besch, Arvind Thiruvengadam, Hemanth K. Kappanna
Institution: West Virginia University, Center for Alternative Fuels, Engines & Emissions
Prepared for: International Council on Clean Transportation
Report date: May 15, 2014 / published by ICCT May 30, 2014
ICCT publication surface:
https://theicct.org/series/dieselgate-behind-the-scandal/page/2/
Historical exact PDF recorded by EPA HERO:
https://theicct.org/sites/default/files/publications/WVU_LDDV_in-use_ICCT_Report_Final_may2014.pdf
EPA HERO record:
https://hero.epa.gov/reference/11581690/
Authority / posture: independent commissioned engineering test report; empirical testing of three selected light-duty diesel vehicles, not regulator adjudication.
May support
PEMS methodology;
route/test conditions;
test-vehicle-specific real-world NOx results;
substantial real-world/lab discrepancy in the tested VW vehicles;
comparison with the tested BMW vehicle.
Important qualification Three test vehicles ≠ fleetwide intentional-cheating proof.
Captureability: MEDIUM-HIGH
Hazard: historical ICCT PDF path may redirect/change; EPA HERO preserves bibliographic identity but not necessarily the report bytes.
Priority: P0
VW-S05 — EPA September 25, 2015 manufacturer-guidance change
Exact source: EPA Update on Recent Volkswagen Announcement + Manufacturer Guidance Letter CD-15-24
Issuer: EPA
Date: September 25, 2015
HTML:
https://www.epa.gov/archive/epa/newsreleases/epa-update-recent-volkswagen-announcement.html
Guidance PDF:
https://www.epa.gov/sites/production/files/2015-10/documents/cd-mfr-guid-ltr-2015-09-25.pdf
Authority / posture: regulator response/process change after Dieselgate discovery.
May support
EPA expanded confirmatory testing to detect potential defeat devices;
discovery altered regulatory test practice.
Counterpedia value A scandal can change the measurement regime itself.
Captureability: HIGH
Priority: P1
VW-S06 — EPA November 2, 2015 second NOV, 3.0L
Exact source: second Notice of Violation and EPA announcement
Issuer: EPA
Date: November 2, 2015
Announcement:
https://www.epa.gov/archive/epa/newsreleases/epa-california-notify-volkswagen-additional-clean-air-act-violations.html
Current documents surface:
https://www.epa.gov/vw/documents-related-volkswagen-violations-model-years-2009-2016
Authority / posture: separate regulator enforcement matter.
May support
then-identified 2014–2016 vehicle scope;
separate discovery from the 2.0L matter.
Important qualification Do not carry 2.0L up to 40× into 3.0L claims.
Captureability: HIGH
Priority: P0
VW-S07 — CARB 3.0L expansion / November 20, 2015
Exact title: Statement from CARB on defeat devices on 3-liter VW, Audi diesel engines
Issuer: CARB
Date: November 20, 2015
Locator:
https://ww2.arb.ca.gov/news/statement-carb-defeat-devices-3-liter-vw-audi-diesel-engines
Authority / posture: regulator statement recording VW/Audi representation that issues extended to all 3.0L diesel engines from model years 2009–2016.
May support
expansion of known vehicle-year scope after the November 2 letter.
Counterpedia value Shows why the "affected vehicle count" is a versioned finding.
Captureability: HIGH
Priority: P1
VW-S08 — CARB January 2016 recall-plan rejection
Exact title: Air Resources Board rejects VW 2-liter diesel recall plan and issues Notice of Violation
Issuer: CARB
Date: January 12, 2016
Locator:
https://ww2.arb.ca.gov/news/air-resources-board-rejects-vw-2-liter-diesel-recall-plan-and-issues-notice-violation
Authority / posture: California regulatory recall/remedy determination.
May support
initial recall proposal was rejected as inadequate;
certification/defeat-device consequences;
remedy was not automatic merely because VW proposed one.
Captureability: HIGH
Priority: P1
VW-S09 — EPA current U.S. violations summary
Exact title: Learn About Volkswagen Violations
Issuer: EPA
Current locator:
https://www.epa.gov/vw/learn-about-volkswagen-violations
Authority / posture: current EPA historical/enforcement synthesis.
Key scope
2.0L model-year 2009–2015: up to 40× standard;
3.0L matter: up to 9× standard.
Captureability: HIGH
Hazard: current webpage is a later synthesis; do not substitute it for the original NOV SourceEdition where exact historical posture matters.
Priority: P1
VW-S10 — Volkswagen AG corporate guilty plea
Exact title: Volkswagen AG Pleads Guilty in Connection with Conspiracy to Cheat U.S. Emissions Tests
Issuer: U.S. Department of Justice
Plea date: March 10, 2017
Locator:
https://www.justice.gov/archives/opa/pr/volkswagen-ag-pleads-guilty-connection-conspiracy-cheat-us-emissions-tests
Case: U.S. v. Volkswagen AG, E.D. Mich.
Authority / posture: corporate criminal guilty plea.
Guilty counts
conspiracy to defraud the United States, commit wire fraud, and violate the Clean Air Act;
obstruction of justice;
importation of merchandise by false statements.
May support
U.S. criminal adjudicative core;
decade-long scheme description through admitted plea record once exact plea agreement is captured;
$2.8B criminal penalty.
Cannot establish
that every board member knowingly participated;
every global jurisdiction's legal conclusion.
Captureability: HIGH for DOJ announcement; exact plea agreement should be captured from DOJ case hub/court.
Priority: P0
VW-S11 — DOJ Volkswagen criminal case hub / plea agreement
Exact source: United States v. Volkswagen / Volkswagen Diesel Engine Vehicle Matters
Issuer: DOJ Criminal Division
Locators:
https://www.justice.gov/criminal/criminal-vns/case/united-states-v-volkswagen
https://www.justice.gov/criminal/criminal-vns/case/volkswagen-diesel-engine-vehicle-matters/overview
Authority / posture: official criminal case repository containing:
criminal information;
plea agreement;
individual indictments/pleas;
updates.
Key U.S. scope Approximately 590,000 diesel vehicles in the U.S. are described in the combined criminal matter.
Captureability: HIGH
Priority: P0
VW-S12 — James Liang guilty plea
Exact title: Volkswagen Engineer Pleads Guilty for His Role in Conspiracy to Cheat U.S. Emissions Tests
Issuer: DOJ
Date: September 9, 2016
Locator:
https://www.justice.gov/archives/opa/pr/volkswagen-engineer-pleads-guilty-his-role-conspiracy-cheat-us-emissions-tests
Authority / posture: individual criminal guilty plea.
Key admitted chronology
Liang admitted that around 2006 he and co-conspirators worked on EA189;
after inability to meet stricter U.S. emissions standards, they designed/implemented software to recognize test vs road operation;
he admitted participation in concealment/misrepresentations.
Cannot establish
knowledge of every senior manager or board member in 2006.
Captureability: HIGH
Priority: P0
VW-S13 — Oliver Schmidt guilty plea and sentence
Exact plea source:Volkswagen Senior Manager Pleads Guilty in Connection With Conspiracy to Cheat U.S. Emissions Tests
https://www.justice.gov/archives/opa/pr/volkswagen-senior-manager-pleads-guilty-connection-conspiracy-cheat-us-emissions-tests
Sentence source:
https://www.justice.gov/archives/opa/pr/volkswagen-senior-manager-sentenced-84-months-prison-role-conspiracy-cheat-us-emissions-tests
Plea: August 4, 2017
Sentence: December 6, 2017
Authority / posture: individual guilty plea and sentencing.
May support
Schmidt's defendant-specific role/knowledge;
later-stage concealment/regulator-deception conduct.
Cannot establish
all-management knowledge.
Captureability: HIGH
Priority: P1
VW-S14 — Martin Winterkorn U.S. indictment
Exact title: Former CEO of Volkswagen AG Charged with Conspiracy and Wire Fraud in Diesel Emissions Scandal
Issuer: DOJ
Date: May 3, 2018
Locator:
https://www.justice.gov/archives/opa/pr/former-ceo-volkswagen-ag-charged-conspiracy-and-wire-fraud-diesel-emissions-scandal
Authority / posture: indictment allegation, not conviction.
Key allegation The indictment alleged Winterkorn was informed of emissions cheating in May 2014 and again in July 2015 and then agreed with others to continue the fraud.
Critical qualification Do not publish:
"A U.S. court found Winterkorn knew in May 2014."
The U.S. source cited here is a charging record.
Captureability: HIGH
Priority: P1
VW-S15 — Volkswagen Group Annual Report 2025, "Diesel issue"
Issuer: Volkswagen AG
Current annual-report page:
https://annualreport2025.volkswagen-group.com/notes/other-disclosures/litigation.html
Authority / posture: current first-party corporate financial/legal disclosure.
Current stated position Volkswagen states:
EA189 software function was developed/implemented from 2006 without knowledge at Board of Management level;
Board members did not learn of development/implementation until summer 2015;
no findings that the unlawful defeat device was disclosed to those preparing 2014 financial statements as the cause of high U.S. NOx;
Volkswagen's legal position includes jurisdiction-specific arguments about unlawfulness.
May support
Volkswagen's current stated legal/factual position;
roughly 11 million EA189-equipped vehicles worldwide as described by Volkswagen.
Cannot establish independently
that Board-level knowledge was absent;
global legal equivalence.
Conflict/qualification Compare with Winterkorn indictment allegations, but do not treat an indictment and corporate annual report as equal adjudicative authority.
Captureability: HIGH
Priority: P1
VW-S16 — EPA Clean Air Act Civil Settlement
Exact title: Volkswagen Clean Air Act Civil Settlement
Issuer: EPA
Locator:
https://www.epa.gov/enforcement/volkswagen-clean-air-act-civil-settlement
Authority / posture: current EPA synthesis of court-approved civil consent-decree settlements.
U.S. scope Approximately 590,000 model-year 2009–2016 diesel vehicles across 2.0L and 3.0L matters.
Settlement structure
vehicle recall/removal or approved modification;
$2.7B mitigation trust for 2.0L;
additional $225M for 3.0L mitigation;
$2B ZEV investment for 2.0L settlement;
$1.45B civil penalty in third partial settlement;
compliance reforms/auditing.
Cannot establish
consumer compensation totals;
criminal penalty;
health-death count.
Captureability: HIGH
Priority: P0
VW-S17 — Exact partial consent-decree family
Issuer/court: U.S. District Court, N.D. Cal. / EPA settlement repository
EPA locator:
https://www.epa.gov/enforcement/third-partial-and-30l-second-partial-and-20l-partial-and-amended-consent-decree
Objects
Amended 2.0L Partial Consent Decree
Authority / posture: entered judicial consent decrees.
Counterpedia value The "Volkswagen settlement" is not one document or one number.
Captureability: HIGH
Hazard: multiple amendments/partial settlements require explicit edition/relationship graph.
Priority: P0
VW-S18 — FTC consumer-deception complaint / 2.0L settlement
Initial charge:FTC Charges Volkswagen Deceived Consumers with Its "Clean Diesel" Campaign
https://www.ftc.gov/news-events/news/press-releases/2016/03/ftc-charges-volkswagen-deceived-consumers-its-clean-diesel-campaign
2.0L settlement announcement:
https://www.ftc.gov/news-events/news/press-releases/2016/06/volkswagen-spend-147-billion-settle-allegations-cheating-emissions-tests-deceiving-customers-20
Authority / posture
complaint = FTC allegations;
stipulated/court-approved order = settlement/remedy.
May support
advertising/consumer-deception theory;
nearly 500,000 2.0L consumers/vehicles in initial U.S. consumer settlement scope;
up to ~$10B consumer buyback/compensation mechanism.
Cannot establish
health damages;
criminal guilt through FTC complaint alone.
Captureability: HIGH
Priority: P1
VW-S19 — FTC 3.0L settlement
Exact title: Federal Trade Commission, Class Action Settlements Require Volkswagen to Repair or Buy Back 3.0 Liter TDI Diesel Vehicles
Issuer: FTC
Date: February 1, 2017
Locator:
https://www.ftc.gov/news-events/news/press-releases/2017/02/federal-trade-commission-class-action-settlements-require-volkswagen-repair-or-buy-back-30-liter-tdi
Authority / posture: consumer settlement/compensation.
May support
different treatment by vehicle generation/model-year;
buyback vs approved repair;
consumer compensation distinctions.
Captureability: HIGH
Priority: P2
VW-S20 — FTC final consumer-redress report
Exact title: In Final Court Summary, FTC Reports Volkswagen Repaid More Than $9.5 Billion To Car Buyers Who Were Deceived by "Clean Diesel" Ad Campaign
Issuer: FTC
Date: July 27, 2020
Locator:
https://www.ftc.gov/news-events/news/press-releases/2020/07/final-court-summary-ftc-reports-volkswagen-repaid-more-95-billion-car-buyers-who-were-deceived-clean
Authority / posture: FTC final status/reporting on consumer-redress program.
May support
$9.5B consumer repayments;
86% of consumers who completed claims process chose buyback/early lease termination;
550,000 deceptively marketed vehicle population across FTC programs.
Cannot establish
total environmental damage;
total all-jurisdiction Dieselgate cost.
Captureability: HIGH
Priority: P1
VW-S21 — Approved emissions modification / configuration change
Exact source family: EPA Documents Related to Volkswagen Violations for Model Years 2009-2016
Locator:
https://www.epa.gov/vw/documents-related-volkswagen-violations-model-years-2009-2016
Example: January 6, 2017 approved modification for Generation 3 model-year 2015 2.0L vehicles
EPA explanatory page:
https://www.epa.gov/vw/january-6-2017-epa-and-california-air-resources-board-approve-remedy-reduce-excess-nox-emissions
Authority / posture: regulator-approved post-scandal modification determination.
May support
exact test-group/generation remedy;
regulator testing/approval of modified emissions configuration;
a repaired vehicle's compliance state differs from its original defeat-device configuration.
Counterpedia value Same VIN/model identity can have different emissions-relevant configurations over time.
Captureability: HIGH
Priority: P1
VW-S22 — EPA current FAQ on excess-emission bounds
Exact title: Frequent Questions about Volkswagen Violations
Issuer: EPA
Locator:
https://www.epa.gov/vw/frequent-questions-about-volkswagen-violations
Authority / posture: current EPA explanatory synthesis.
Key measurement scope
2.0L: NOx emission levels 10–40 times higher than emission standards;
3.0L: up to 9 times standards;
EPA states collective emissions contribute to air pollution and health effects; it does not frame one violating car as an acute passenger-compartment health threat.
Captureability: HIGH
Priority: P1
VW-S23 — Barrett et al. modeled U.S. public-health impact
Exact title: Impact of the Volkswagen emissions control defeat device on US public health
Authors: Steven R. H. Barrett et al.
Journal: Environmental Research Letters 10 (2015), 114005
DOI: 10.1088/1748-9326/10/11/114005
Authoritative bibliographic/abstract sources:
https://hero.epa.gov/reference/3074890/
https://www.hks.harvard.edu/publications/impact-volkswagen-emissions-control-defeat-device-us-public-health
Authority / posture: peer-reviewed model-based health-impact estimate, not regulator adjudication or individual medical causation.
Key estimate
59 premature deaths central estimate over 2008–2015 U.S. sales period;
95% CI approximately 10–150;
about $450M modeled social cost;
counterfactual recall scenarios.
Method chain estimated excess NOx → spatial emissions distribution → atmospheric exposure model → concentration-response functions → modeled mortality/morbidity.
Cannot establish
identities of 59 specific people;
that exactly 59 deaths occurred;
causation for an individual death.
Captureability: HIGH bibliographic/repository; acquire exact article bytes where license/origin permits.
Priority: P0 for model/inference discipline.
VW-S24 — Holland et al./PMC alternative modeled health-impact analysis
Exact title: Public Health Impact and Economic Costs of Volkswagen's Lack of Compliance with the United States' Emission Standards
Peer-reviewed full text:
https://pmc.ncbi.nlm.nih.gov/articles/PMC5036724/
Authority / posture: peer-reviewed model-based estimate.
Estimate range Using scenario assumptions for annual excess NOx:
modeled premature deaths ranged roughly 5–50 for one year of operation, depending on emissions and mortality assumptions;
other morbidity/economic outputs varied materially across assumptions.
Counterpedia value Different health-impact studies can yield different numbers because they model:
emissions magnitude;
exposure;
concentration-response;
time horizon;
fleet assumptions.
Captureability: HIGH
Priority: P1
VW-S25 — Wikipedia comparison surface — OPTIONAL
Title: Volkswagen emissions scandal
Role: familiar secondary comparison surface only
Capture rule: pin exact oldid.
Authority / posture: secondary collaborative narrative.
Priority: OPTIONAL
C. RECORD-CONTRADICTION / QUALIFICATION MAP
C-1 — Real-world excess emissions vs intentional defeat-device finding
WVU/ICCT 2014
Measured substantial real-world NOx exceedances in two tested VW diesels.
EPA/CARB 2015
Regulatory testing/investigation established software behavior that distinguished test conditions and normal use.
Volkswagen AG 2017
Corporate guilty plea established criminal responsibility for a U.S. conspiracy involving emissions-test cheating.
Counterpedia rule Do not retroactively make the 2014 study say:
"WVU proved VW had a defeat device."
The study established empirical discrepancy. The intentional mechanism was established later.
C-2 — 40× vs 35× vs 20× vs 9×
Different sources report different bounds:
WVU test-vehicle/route results: particular tested VW vehicles showed different multipliers depending on vehicle/route.
EPA U.S. 2.0L synthesis: 10–40× standards.
EPA 3.0L matter: up to 9× standards.
These numbers differ because:
vehicle/engine;
route/drive cycle;
numerator;
regulator standard;
measurement context.
A page should never say:
"Volkswagens emitted exactly 40 times the legal limit."
without vehicle/configuration/test scope.
C-3 — 482k / ~500k / ~550k / ~590k / ~11m are not contradictory
Different counts refer to different scopes and dates.
initial U.S. 2.0L enforcement scope: roughly 482k / nearly 500k;
FTC combined consumer programs: >550k vehicles;
DOJ/EPA combined U.S. 2.0L + 3.0L matter: approximately 590k;
Volkswagen global EA189 statement: roughly 11m vehicles worldwide.
Counterpedia should attach: jurisdiction
enginemodel yearsdate of knowledge
to every count.
C-4 — Corporate guilt ≠ every executive's knowledge
Volkswagen AG pleaded guilty to a corporate conspiracy.
Liang and Schmidt separately pleaded guilty.
But claims such as:
"Volkswagen's entire board knew since 2006"
require actor-specific evidence.
Volkswagen's current annual report says the software was developed/implemented from 2006 without Board-level knowledge and that Board members learned in summer 2015.
The U.S. indictment of former CEO Martin Winterkorn alleged he was informed in May 2014 and July 2015.
These source postures are not symmetrical:
corporate annual report = interested first-party position;
indictment = government allegation;
neither, standing alone, is a conviction establishing exact Board knowledge timeline.
C-5 — Company admitted defeat-device scheme ≠ every individual defendant convicted
Volkswagen corporate plea is adjudicated.
Liang and Schmidt have individual guilty pleas.
Other named executives were charged/indicted in U.S. proceedings with different procedural outcomes/availability.
Counterpedia must preserve individual disposition rather than inheriting corporate guilt to every person.
C-6 — Clean Air Act violation vs consumer deception
EPA/DOJ environmental case:
certification;
defeat devices;
excess NOx;
mitigation;
civil penalties.
FTC consumer case:
advertising/marketing;
environmental claims;
resale/value representation;
consumer compensation.
One set of facts can support multiple legal theories, but settlement amounts and elements are not interchangeable.
C-7 — Consumer compensation vs pollution mitigation
U.S. remedy structure includes:
up to ~$10B consumer compensation/buybacks for 2.0L;
$9.5B later reported consumer repayments;
$2.7B + $225M mitigation trust;
$2B ZEV investment;
$1.45B civil CAA penalty;
$2.8B criminal penalty.
These are different legal/remedial purposes.
Do not add them and label the sum:
"the amount of environmental damage."
C-8 — Defeat-device vehicle before remedy vs after approved modification
EPA/CARB approved specific emissions modifications for particular generations/test groups after independent testing.
Thus: 2015 VW Golf TDI with original defeat-device calibration and same vehicle after approved emissions modification
can be the same physical vehicle identity but different regulated emissions configuration states.
Counterpedia should time-bind emissions claims.
C-9 — "Clean Diesel" was deceptive ≠ diesel technology is universally incapable of low NOx
The original WVU project included a BMW comparison vehicle whose real-world performance was materially different from the tested VW vehicles.
Therefore Dieselgate establishes Volkswagen-specific cheating/noncompliance in the affected corpus; it does not by itself establish:
all diesel technology is incapable of regulatory compliance.
C-10 — 59 deaths vs 5–50 deaths vs "no confirmed individual health case"
Peer-reviewed studies estimate health impacts using models.
Barrett et al.:
central estimate 59 premature deaths;
broad confidence interval.
Holland et al.:
scenario-dependent one-year range with materially different mortality outputs.
EPA:
describes population-level adverse health risks from collective NOx pollution;
says emissions from a single violating car are not framed as an acute passenger-compartment hazard.
No one of those means:
"Regulators identified exactly 59 people killed by Volkswagen cars."
Modelled attributable mortality is a population-statistical proposition.
C-11 — Emissions estimate → health estimate has multiple transformations
A health-impact model needs:
affected fleet;
miles driven;
excess NOx per mile;
geographic allocation;
atmospheric chemistry/dispersion;
population exposure;
concentration-response relationship;
counterfactual baseline.
The output is not a direct sensor reading.
This should be visible in the evidence graph.
C-12 — U.S. unlawfulness ≠ automatic global legal equivalence
Volkswagen's current annual report explicitly preserves its legal position that aspects of the relevant software have jurisdiction-specific legal treatment.
Counterpedia need not decide the full European legal issue in this U.S. demo.
It should refuse:
"Because EPA called it illegal, every one of the roughly 11m global vehicles was adjudicated illegal under every country's law."
That requires jurisdiction-specific authority.
D. FIVE-CLAIM DEMONSTRATION MATRIX — PROVISIONAL ONLY
C1 — Settled core event/conduct fact
Provisional wording
Volkswagen AG used prohibited defeat-device software in affected U.S. diesel vehicles so that emissions controls behaved differently during regulatory testing than during normal driving, and Volkswagen AG later pleaded guilty to federal felony charges arising from the U.S. emissions-cheating scheme.
Candidate sources
Can establish
defeat-device conduct in affected U.S. vehicles;
corporate criminal plea;
U.S. jurisdictional scope.
Cannot establish
every VW diesel model globally;
every individual's knowledge.
Likely posture
regulator finding + corporate criminal adjudication.
Qualification Separate 2.0L and 3.0L technical scopes where specifics matter.
C2 — Measurement / structural fact
Provisional wording
EPA reports that affected U.S. 2.0L vehicles emitted NOx at roughly 10–40 times the applicable standard under relevant normal-driving conditions, while the affected 3.0L vehicles were reported at up to about nine times the standard; these are fleet/configuration bounds, not a universal 40× value for every affected car.
Candidate sources
Can establish
regulator-stated bounds;
engine-family distinction.
Cannot establish
exactly 40× for every car/mile;
one average across all affected models.
Likely posture
regulator measurement synthesis.
Refusal branch
"Every cheating VW emitted forty times the legal limit whenever driven."
Not supported.
Strengthening A vehicle-specific claim requires exact test-group/VIN configuration + measurement protocol.
C3 — Discovery / action chronology
Provisional wording
Independent on-road testing commissioned by ICCT and performed by WVU found unexpectedly high real-world NOx emissions in tested Volkswagen diesels; subsequent CARB/EPA investigation identified defeat-device behavior, EPA issued its first Notice of Violation on September 18, 2015, and regulators later expanded testing and the affected U.S. vehicle scope.
Candidate sources
Can establish
research-to-regulator chronology;
distinction between discrepancy discovery and intentional-mechanism determination.
Cannot establish
that WVU itself proved criminal intent.
Likely posture
research observation → regulatory investigation → enforcement.
C4 — Carefully bounded public-health proposition
Provisional wording
The affected vehicles emitted excess NOx, a pollutant precursor associated with population-level respiratory, cardiovascular, ozone and particulate-matter harms; peer-reviewed studies estimated premature mortality attributable to the excess U.S. emissions, but the numerical death estimates are model-dependent rather than counts of individually identified deaths.
Candidate sources
Can establish
excess NOx population health concern;
published model estimates and uncertainty.
Cannot establish
exact individual victim causation;
a single uncontested death count.
Likely posture
regulator health-risk statement + model-dependent scientific inference.
Mandatory UI Display: observed/estimated emissions → air-quality model → exposure model → concentration-response → estimated attributable outcomes.
C5 — Contested / commonly compressed proposition
Popular formulation
"Volkswagen's top leadership knowingly ran the defeat-device scheme for a decade."
What the record strongly establishes
the U.S. scheme began around 2006 within the engineering/conspirator record;
Volkswagen AG later pleaded guilty;
James Liang admitted participation beginning during EA189 development;
Oliver Schmidt later pleaded guilty to his own role;
U.S. prosecutors charged additional executives.
What the currently located record does not establish at the proposed universal scope
every Board member knew from 2006;
all senior executives shared the same knowledge date;
an adjudicated single timeline for the entire Board.
Competing/posture-specific sources
Volkswagen 2025 Annual Report says development/implementation occurred without Board-level knowledge and Board members learned in summer 2015.
U.S. indictment alleged Martin Winterkorn was informed in May 2014 and again in July 2015.
Why this remains nonbinary The corporate criminal core is settled; the actor-by-actor knowledge graph is not reducible to one company-wide date.
Stronger admissible wording
Volkswagen AG admitted a long-running U.S. emissions-cheating conspiracy, and individual employees admitted participation; claims about when particular senior executives or Board members learned of the defeat device require person-specific evidence and procedural posture.
Reopening condition For a named person:
authenticated contemporaneous communication/minutes/testimony;
guilty plea, trial finding or other adjudication;
exact date and content of information received.
A corporate plea cannot alone fill the individual knowledge edge.
E. BEST "WHY NOT?" SPECIMEN
Popular formulation
"Volkswagen killed 59 Americans with Dieselgate."
Why Counterpedia should refuse that formulation at face value
The 59 figure is real and useful—but it is a model output, not a list of 59 adjudicated deaths.
Barrett et al. modelled:
excess NOx estimate → geographic emissions → atmospheric PM/ozone exposure → concentration-response → 59 central estimated premature deaths
with a wide confidence interval, roughly 10–150.
Another peer-reviewed analysis used different assumptions/scenarios and produced a different range.
So:
A peer-reviewed model estimated about 59 U.S. premature deaths attributable to excess emissions, with substantial uncertainty.
is supportable.
Volkswagen was proven to have killed exactly 59 named Americans.
is not.
Reopening condition
An individual-causation claim would require person-specific exposure and medical/epidemiologic evidence capable of supporting that causal inference.
A population attributable-risk model cannot provide individual identity by itself.
F. SECOND "WHY NOT?" — EVERY CAR WAS 40×
Popular formulation
"The cheating Volkswagens polluted 40 times over the legal limit."
This is useful shorthand only if explicitly scoped.
The record distinguishes:
specific WVU tested vehicles/routes;
2.0L EPA range around 10–40×;
3.0L up to ~9×;
later approved modified configurations.
The governed version needs: engine family
model year/test groupconfigurationtest/road conditionstatistic/range.
G. SOURCE-PAGE / REVERSE-WIKIPEDIA VALUE
1. VW-S04 — WVU/ICCT 2014 report
Why ideal
This source shows an observation becoming a scandal without retroactively changing the observation.
Reverse traversal:
PEMS measurements → tested Jetta/Passat/BMW → unexpected VW real-world discrepancy → CARB/EPA investigation → defeat-device determination → later criminal plea
The Source Page can explicitly say:
this report measured the symptom; it did not itself adjudicate the intent/mechanism.
2. VW-S01 — September 2015 EPA NOV
Why ideal
One click can show:
exact 2.0L vehicle/test-group scope;
regulator allegations;
original
up to 40×context;later criminal/civil sources that strengthen or supersede posture.
This is an excellent allegation → adjudication lineage Source Page.
3. VW-S10/VW-S11 — Volkswagen corporate plea / plea agreement
Why ideal
This is the settled-core anchor.
Reverse traversal: corporate plea → exact admitted conspiracy conduct → U.S. affected fleet → obstruction/importation counts → individual employee proceedings → claims it does not establish about every executive.
4. VW-S23 — Barrett health-impact model
Why ideal
This is a model transparency Source Page.
Reverse traversal: 59 → confidence interval → fleet assumptions → excess emissions assumptions → atmospheric model → concentration-response → every article claim quoting "59 deaths."
Users can see the transformations under a headline statistic.
H. 30–60 SECOND DEMO MOMENT
Open Dieselgate with three statements:
SETTLED
Volkswagen installed defeat-device software in affected U.S. diesels and Volkswagen AG pleaded guilty to a federal emissions-cheating conspiracy.
COMMON NUMBER
The cars emitted 40× the legal limit.
COMMON CONSEQUENCE
Dieselgate killed 59 people.
Click Why this page?
Counterpedia resolves them:
DEFEAT DEVICE
High-confidence / adjudicative core
40×
QUALIFIED
2.0L: EPA reports 10–40× range;
3.0L: up to 9×;
not every vehicle/mile;
configuration and test condition matter.
59 DEATHS
MODEL-DEPENDENT
central peer-reviewed estimate;
roughly 10–150 uncertainty interval;
atmospheric/exposure/concentration-response transformation;
not 59 named adjudicated victims.
Then click:
Who knew?
Counterpedia does not say "Volkswagen knew" as if a corporation has one timestamp.
It opens:
Liang → admitted timeline;
Schmidt → admitted timeline;
Winterkorn → indictment allegations;
Volkswagen Board → company's current first-party position.
The intellectual contrast is:
The core fraud can be settled while vehicle scope, emission magnitude, health impact and actor-specific knowledge remain different evidence problems.
I. CAPTURE PRIORITY
P0 — essential
VW-S10/VW-S11 Volkswagen criminal information + exact plea agreement + judgment
VW-S12 Liang plea agreement
VW-S21 at least one exact approved emissions-modification artifact
VW-S23 Barrett et al. article / authoritative exact edition
P1 — important qualification / contradiction
VW-S05 Sept. 25 enhanced-testing guidance
VW-S07 3.0L scope expansion
VW-S08 recall-plan rejection
VW-S13 Schmidt plea/judgment
VW-S14 Winterkorn indictment
VW-S15 Volkswagen 2025 annual-report legal position
VW-S18 FTC complaint + 2.0L order
VW-S20 FTC final redress report
VW-S24 alternative peer-reviewed health-impact study
P2
VW-S19 3.0L consumer settlement
subsequent PEMS compliance/auditor reports
specific vehicle/test-group certification records where a narrow emissions claim is pursued
OPTIONAL
VW-S25 Wikipedia comparison edition pinned by oldid
J. INTEGRITY FLAGS
1. NOV posture is not guilty-plea posture
Preserve enforcement chronology.
2. WVU report is observation, not intent finding
Mandatory qualification.
3. Do not generalize 40× to all affected vehicles
Engine/configuration/test conditions matter.
4. Do not copy 2.0L multiplier to 3.0L
Different EPA bounds.
5. Vehicle counts need jurisdiction/engine/date
482k / 500k / 550k / 590k / 11m are different scope objects.
6. Corporate guilt is not universal individual guilt
Person-specific disposition required.
7. Indictment is not conviction
Especially Winterkorn/other charged executives.
8. Volkswagen annual report is first-party posture
Do not treat its Board-knowledge statement as neutral adjudication.
9. U.S. legal conclusion cannot be silently globalized
Jurisdiction matters.
10. Consumer remedy ≠ environmental remedy
Never merge settlement dollars by numeric proximity.
11. Criminal penalty ≠ civil penalty ≠ mitigation trust ≠ ZEV investment
Label remedy type.
12. Health estimates are model outputs
Carry model, assumptions and uncertainty.
13. "59 deaths" does not identify 59 persons
No individual-causation laundering.
14. NOx → health is a transformation chain
Do not render model outputs as direct tailpipe measurements.
15. Approved modifications create configuration history
Same VIN can have pre-remedy and post-remedy emissions state.
16. EPA current pages are later syntheses
For historical claims, prefer original SourceEdition.
17. Consent decrees can be amended/partial
Capture exact entered edition.
18. Court/agency hosting does not change authorship of attached corporate artifacts
Preserve source posture.
19. ICCT historical PDF path may drift
Capture exact bytes; don't infer equivalence from title alone.
20. "Fully compensate" in FTC language has remedy-program scope
It is not proof that every conceivable consumer or consequential loss was economically erased.
K. DIESELGATE CAPTURE DISPATCH SHAPE
For VW-CAP1, acquire in this order:
criminal information
plea agreement
judgment
VW-S12 Liang:
indictment
plea agreement
judgment
VW-S13 Schmidt:
information
plea
judgment
VW-S14 Winterkorn indictment as allegation-only source
VW-S18 FTC complaint + exact stipulated 2.0L order
VW-S19 3.0L consumer order
VW-S20 final consumer-redress report
VW-S21 exact approved emissions modification(s), including test-group scope
VW-S15 Volkswagen 2025 Annual Report exact page/PDF edition
VW-S23 Barrett et al.
VW-S24 alternative health-impact paper
optional later PEMS/auditor reports
CAP1 output remains intentionally boring:
requested locator
final locator
HTTP result
media type
exact bytes/digest
capture observation
authoring institution
source-owned date
engine family
model years/test group
original vs modified configuration
measured value/range
denominator/standard
legal posture
individual/corporate actor
model assumptions/uncertainty for health sources
document amendment/version
captured / refused / failed
no final standing
no article prose
L. PROVISIONAL DIESELGATE THESIS FOR REVIEW
Dieselgate should not be demoed as:
"Did Volkswagen cheat? Yes, verified."
That core question is too easy.
The stronger Counterpedia thesis is:
A scandal can have an unusually settled core while nearly every surrounding headline still requires scope.
The system should be comfortable saying all of these at once:
SETTLED CORE
affected U.S. vehicles used defeat-device software;
Volkswagen AG pleaded guilty to a long-running U.S. emissions-cheating conspiracy.
SCOPE-BOUND
affected vehicle count depends on engine, jurisdiction, and date;
2.0L and 3.0L emissions multipliers differ;
approved remedies change vehicle configuration.
ACTOR-SPECIFIC
some employees pleaded guilty;
other executives were charged;
Board-level knowledge timing cannot be inherited from corporate guilt.
MODEL-DEPENDENT
premature-death estimates depend on emissions, atmospheric, exposure, response and fleet assumptions.
The key reveal is:
"40×" can be true without every car emitting 40×, and "59 deaths" can be a serious peer-reviewed result without there being 59 individually proven Dieselgate death cases.
Dieselgate therefore exercises:
settled core vs disputed scope;
regulatory allegation vs criminal admission;
independent measurement vs intent finding;
engine/model-year/jurisdiction denominators;
vehicle configuration history;
actor-specific organizational knowledge;
indictment vs guilty plea;
consumer vs environmental vs criminal remedies;
population health modelling vs individual causation;
and numeric claims whose meaning depends on the transformation chain beneath them.
The Dieselgate invariant is:
A settled scandal does not license imprecise surrounding claims. Strong evidence should make Counterpedia more precise, not less.