# Research dossier

Run `41c563cb-fda0-42e6-9fed-56983e8a3086`.

Asked as:

```
Project MKUltra
```

This document is a **method transcript, not a verdict**. It records what was
asked, where this run looked, what it read, how each passage bore on each
competing explanation, and what it could not settle — in that order, so a
reader can perform the adjudication instead of being handed one. Nothing
below asserts that the run's conclusions are correct.

### TL;DR — provisional read

This is a navigation aid, not a new synthesis. It restates the first three
asserted findings in the run's own synthesis order and points to their audit
trail in §§6–13; it does not rank them by the ordinal score matrix.

#### Current read

- `fnd_36e0d278d5795134_00` [observed; computed confidence: unknown] The ledger favours H1 over both alternatives at the level it actually speaks to: unwitting administration of MKULTRA substances to U.S. citizens was a standing programme element whose continuation was formally deliberated in the office of the Deputy Director of Central Intelligence, with named senior officials (Gottlieb and Helms) arguing for continuation and others (Earman, Carter, Kirkpatrick) opposing, and with the participants agreeing the question might have to be escalated to the DCI for decision (ev_88177732ca07_01, ev_88177732ca07_02, ev_88177732ca07_05, ev_88177732ca07_06). A practice that requires DDCI- and DCI-level adjudication and a formal 'freeze' is a managed policy, not an accident of oversight. *Why that level:* 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin.
- `fnd_71b81992bd8ad0f3_01` [inferred; computed confidence: unknown] H3's premise that the absence of ethical review reflected merely undeveloped contemporaneous norms is undercut by the grounds on which restraint was actually argued inside the Agency: unwitting testing on foreign nationals 'had been ruled out' because station chiefs judged it 'too dangerous' and lacked 'controlled facilities' (ev_88177732ca07_03), the domestic safehouse arrangements were criticised internally as slipshod (ev_88177732ca07_04), and continuation was framed as a 'problem' warranting director-level sign-off (ev_88177732ca07_05). The recorded calculus is operational risk and exposure, and the recorded objection is internal — which means an ethical objection existed within the institution at the time, contrary to a 'no norms yet' account. *Why that level:* 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin.
- `fnd_4b5395942e8f1a8a_03` [inferred; computed confidence: unknown] H2's core quantitative claim is untouched rather than refuted by this ledger, and this is the sharpest limit on the whole reading: every item is drawn from one policy-level deliberation and its 1975 cover memo (ev_88177732ca07_01, ev_88177732ca07_02, ev_88177732ca07_05, ev_88177732ca07_06). There is no subject-level record, no consent documentation, no subproject inventory and hence no denominator, so the ledger cannot say what proportion of the ~150 subprojects involved covert administration versus witting volunteers, contract university science, or animal work. *Why that level:* 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin.

#### In plain language

> Project MKUltra secretly gave drugs to Americans without their agreement. Senior CIA officials argued about whether to continue this practice.

*Written by a language model from the findings below — the only prose in this
document a model composed. Measured reading grade: 11.3 (Flesch-Kincaid). This is above the eighth-grade level the note aims for; it is printed as written rather than withheld. It restates the findings; it is not evidence, nothing
cites it, and no stage of the research could read it.*

#### Evidence score

**6 findings asserted, unknown confidence — 14 / 100.**
0 independent origin(s) behind 6 asserted finding(s); 7 recorded stress attack(s). Components summed to 29; 1 source(s) could not be traced to an origin, which caps the unknown band.

Components: corroboration 0/40 · provenance 0/25 · falsification 14/20 · saturation 15/15.

**Numeric probability: not estimated.** The score above is a stated transform
of the counts printed beside it, bounded so it can never outrank the computed
confidence level it reports. It is not a probability: this run has no
calibrated priors or likelihood model, and the `++`, `+`, `0`, `-`, `--`
matrix is ordinal and cannot honestly be converted into a percentage. A reader
who rejects the weighting still has every count it was made from.

#### Strongest directly cited passages

Selected only from citations attached to the findings above: traced origin
first, then the evidence hierarchy in force at render time, then ledger order.
A higher publisher tier does not repair untraced provenance.

- `ev_88177732ca07_01` — origin untraced; tier: government institutional; John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive: This memo records a meeting held in the office of Deputy Director of Central Intelligence Gen. Marshall Carter to settle the one major point of disagreement among CIA officials over the inspector general’s MKULTRA recommendations: whether to continue with the testing of MKULTRA substances on unwitting U.S. citizens.
- `ev_88177732ca07_02` — origin untraced; tier: government institutional; John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive: Both Gottlieb and Helms “argued for the continuation of unwitting testing,” while Earman, Carter and Kirkpatrick disagreed.
- `ev_88177732ca07_03` — origin untraced; tier: government institutional; John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive: Carter was concerned with the “unwitting aspect,” and a discussion ensued “on the possibility of unwitting test on foreign nationals,” which “had been ruled out” due to opposition from “senior chiefs of stations” as “too dangerous” and who said they lacked “controlled facilities.”

#### What would change or overturn this read

These are the run's recorded stress conditions, not guarantees that one new
fact alone would settle the whole question. See §13 for every attack.
- `stp_2e9d970e515e43db_00` [causal_overreach] targeting `fnd_36e0d278d5795134_00` — High-level DDCI deliberation and an executive freeze do not establish that unwitting administration was an established, managed policy program-wide; executive intervention may have been an ad hoc containment response to rogue or compartmentalized activities brought to light by the IG. **Recorded failure condition:** Contemporaneous records demonstrate that the DDCI meeting was convened strictly as crisis containment for isolated, unauthorized safehouse experiments rather than a managed review of an authorized Agency-wide policy.
- `stp_a7fee67587392a95_01` [alternative_explanation] targeting `fnd_71b81992bd8ad0f3_01` — Internal objections grounded in operational danger, lack of controlled facilities, and slipshod safehouses reflect tactical risk management and fear of public exposure, not contemporaneous biomedical ethical norms. **Recorded failure condition:** Archival evidence demonstrates that internal pushback was motivated exclusively by security vulnerability, diplomatic blowback, and legal exposure without reference to ethical or human-subjects protection principles.
- `stp_87495abc063099c9_02` [source_dependence] targeting `fnd_4b5395942e8f1a8a_03` — Concluding that H2's quantitative distribution cannot be evaluated relies on the ledger's single executive memo chain, ignoring external documentary streams such as surviving subproject financial ledgers or institutional records. **Recorded failure condition:** Financial disbursement records, contract files, or institutional archives provide a verifiable denominator showing what proportion of subprojects engaged in non-consensual testing versus standard academic research.

## 1. Research Question

To what extent did Project MKUltra's documented human subjects research activities involve non-consensual experimentation, and what were the program's stated objectives, methods, and institutional oversight failures?

## 2. Scope

- **Population:** Human subjects (witting and unwitting) involved in CIA-sponsored or CIA-funded behavioral, psychological, and pharmacological research; researchers, contractors, and institutions receiving CIA funding under this and related subprojects
- **Geography:** Primarily United States, with some subprojects and activities extending to Canada and other allied countries where CIA-funded research was conducted
- **Period:** 1953 to 1973 (official program duration as documented), with relevant precursor programs (e.g., Project BLUEBIRD, Project ARTICHOKE) from 1950 and aftermath including 1975-1977 Congressional investigations and document releases

### Out of scope

- Detailed technical/scientific evaluation of the validity of research findings produced under MKUltra
- Broader Cold War intelligence history not directly tied to MKUltra's human experimentation activities
- Post-1977 CIA behavioral research programs not derived from or continuous with MKUltra
- Civil litigation outcomes and individual compensation claims by test subjects, except as they bear on documenting scope of activities
- Popular culture depictions and fictionalized treatments of MKUltra

### Ambiguities the question recorded

- The full scope of MKUltra remains uncertain because most records were ordered destroyed in 1973; conclusions may differ depending on whether the question is scoped to documented/surviving records only or presumed full extent
- Whether 'MKUltra' should be treated as coextensive with the broader family of related CIA programs (e.g., MKDELTA, MKNAOMI, Project ARTICHOKE) or as a narrower, distinct program affects population and period boundaries
- The number of subprojects and total human subjects affected varies across sources (Church Committee, Rockefeller Commission, subsequent FOIA releases), which would change quantitative answers

### Assumptions the question recorded

- This scope treats 'documented' evidence as that surfacing from the 1977 Senate Church Committee hearings, the 1975 Rockefeley Commission, and the roughly 20,000 surviving financial documents located after the 1973 destruction order, rather than speculative or unverified claims
- Related precursor and successor programs (BLUEBIRD, ARTICHOKE) are treated as contextual background rather than within the core population unless explicitly designated as MKUltra subprojects

## 3. Definitions

- MKUltra: the CIA's classified program of human experimentation, officially initiated in 1953, involving behavioral modification, interrogation techniques, and drug testing (notably LSD), organized into numbered subprojects
- Non-consensual experimentation: research conducted without informed consent, including covert drug administration and psychological manipulation
- Institutional oversight failure: absence or circumvention of ethical review, legal authorization, or accountability mechanisms governing the research
- Subproject: a discrete, separately funded research activity conducted under the MKUltra umbrella, often at universities, hospitals, prisons, or through front organizations

## 4. Methodology

### Evidence standard

The hierarchy below is **the one in force at render time**. It is a module
constant, not per-run state — no field on this run stores the ladder its
sources were ranked against — so this document can say which ordering is in
force as these words are written, and no more. Re-rendering an older run
after the constant changes would show that run the newer ladder.

| Rank | Tier | What qualifies |
|---|---|---|
| 1 | primary_data | Primary experimental or observational data: the dataset, trial record, or measurement itself, not a description of it. |
| 2 | systematic_review | A systematic review or meta-analysis that pools and appraises primary studies against a stated method. |
| 3 | peer_reviewed | A single peer-reviewed study, not yet pooled or independently replicated. |
| 4 | government_institutional | A government or institutional dataset or report published by the body that collected or regulates the data. |
| 5 | technical_report | A technical report describing methodology and results outside the peer-review process. |
| 6 | expert_analysis | Analysis by a named subject-matter expert, not independently reviewed. |
| 7 | quality_journalism | Reporting from a publication with a named byline, editorial process, and correction record. |
| 8 | industry_commentary | Commentary from an interested party writing about its own field. |
| 9 | anecdotal | A single first-hand account, not generalized beyond itself. |
| 10 | unsupported | Social media, or a claim with no stated basis at all. |

### How absence is written in this document

Three different absences look identical on a page. This document keeps them
apart deliberately, because a measured zero and an absent measurement are not
the same silence.

- *"... has not run"* appears only where a field distinguishes it: the
  question, the landscape, the search strategy, the falsification report and
  the stress test are each absent until their stage produces one.
- *"This run holds no ..."* appears for every collection. This is weaker on
  purpose: **an empty collection cannot say** whether the stage that would
  have filled it ever ran. One empty value stands for two different facts,
  and nothing in this run's state separates them. That is a defect of the
  state, stated once, here.
- *"This run does not record ..."* appears where there is no field at all — a
  fact this pipeline never captured, rather than one it captured as empty.

### Searches

| Query | Kind | Site | Compiled from |
|---|---|---|---|
| Project MKUltra human subjects research non-consensual experimentation institutional oversight failures | anchor | — | MKUltra |
| CIA MK-ULTRA mind control behavioral engineering experiments | synonym_expansion | — | MKUltra, behavior modification |
| CIA human experimentation without informed consent LSD testing subjects | synonym_expansion | — | human subjects research, informed consent, LSD experimentation |
| MKUltra numbered subprojects front organizations funding conduits universities hospitals prisons | synonym_expansion | — | subproject, front organizations |
| Church Committee final report CIA human experimentation | site_scoped | archive.org | Church Committee |
| MKUltra declassified records | site_scoped | nsarchive.gwu.edu | National Security Archive, Freedom of Information Act release |
| Rockefeller Commission report CIA activities within the United States human experimentation | review_hunt | — | Rockefeller Commission report (1975) |
| historiographical review declassified MKUltra financial and subproject documents | review_hunt | — | Meta-analyses or historiographical reviews of declassified MKUltra financial and subproject documents |
| Institute of Medicine National Academies report human subjects protections Cold War abuses | review_hunt | — | Institute of Medicine or National Academies reports on human subjects protections arising from Cold War-era abuses |
| bioethics literature review informed consent failures government-sponsored research Nuremberg Code violations | review_hunt | — | Legal and bioethics literature reviews on informed consent failures in government-sponsored research, biomedical ethics violations |

### Subquestions

- `SQ1` — What were the officially stated objectives and goals of Project MKUltra as documented in surviving CIA records and Congressional testimony?
- `SQ2` — What specific research methods and techniques (e.g., drug administration, sensory deprivation, hypnosis, electroshock) were used across documented MKUltra subprojects?
- `SQ3` — How many subprojects and institutions (universities, hospitals, prisons, front organizations) are documented as receiving CIA funding under MKUltra, based on surviving financial records?
- `SQ4` — What evidence exists that human subjects were experimented on without their knowledge or consent, and what proportion of documented activities involved such non-consensual experimentation versus witting participation?
- `SQ5` — What specific cases or subprojects are documented as involving covert administration of drugs (e.g., LSD) to unwitting subjects, and what were the reported consequences for those subjects?
- `SQ6` — What formal ethical review, legal authorization, or informed consent procedures (if any) governed MKUltra research activities at the time they were conducted?
- `SQ7` — In what ways did MKUltra activities circumvent or violate existing legal and ethical standards (e.g., the Nuremberg Code) applicable to human subjects research during this period?
- `SQ8` — What institutional oversight mechanisms existed within the CIA and associated agencies to monitor MKUltra, and what failures or gaps in oversight have been documented?
- `SQ9` — How did the 1973 destruction of records affect the ability of subsequent investigators (Rockefeller Commission, Church Committee) to document the full scope of MKUltra's human subjects research?
- `SQ10` — What findings and conclusions did the Rockefeller Commission (1975) and Church Committee (1977) reach regarding the extent of non-consensual experimentation and oversight failures in MKUltra?
- `SQ11` — Where do sources (Church Committee, Rockefeller Commission, FOIA-released documents) disagree on the scope, number of subjects, or extent of non-consensual experimentation under MKUltra?
- `SQ12` — What role did precursor programs (BLUEBIRD, ARTICHOKE) play in establishing methods or objectives later continued under MKUltra, insofar as they are documented as directly linked to MKUltra subprojects?
- `SQ13` — What documented evidence exists of extension of MKUltra-related human experimentation activities to Canada or other allied countries, and how did oversight arrangements differ there?
- `SQ14` — What remains unknown or undocumented about the full extent of MKUltra's non-consensual experimentation due to the incompleteness of surviving records?
- `SQ15` — What accountability measures, if any, were taken against individuals or institutions responsible for oversight failures in MKUltra, as documented in Congressional investigations?

### Retrieval rounds

- Round 1, the initial retrieval: 4 snapshot(s).
- Round 2, the falsification counter-search: 0 snapshot(s).

## 5. Current State of Knowledge

**This is not what the field believes.** It is where this run looked,
decided **before any source was read** — vocabulary to search with, places
to search in, and the kinds of review to hunt for. It contains no claim about
the world, and **no source backs any of it**. Nothing downstream reads it as
a claim either, so the worst a bad map can do is search badly.

### Terminology searched

- `MKUltra` — also written `MK-ULTRA`, `MK Ultra`, `Project MKULTRA`
- `Project BLUEBIRD` — also written `Operation Bluebird`
- `Project ARTICHOKE` — also written `Operation Artichoke`
- `MKDELTA` — also written `MK-DELTA`
- `MKNAOMI` — also written `MK-NAOMI`
- `human subjects research` — also written `human experimentation`, `human subject experimentation`
- `informed consent` — also written `voluntary consent`, `subject consent`
- `behavior modification` — also written `behavioral control`, `mind control`, `behavioral engineering`
- `interrogation techniques` — also written `coercive interrogation`, `enhanced interrogation (historical usage)`
- `chemical and biological research` — also written `psychochemical research`, `pharmacological research`
- `LSD experimentation` — also written `lysergic acid diethylamide testing`, `hallucinogen research`
- `Church Committee` — also written `Senate Select Committee to Study Governmental Operations with Respect to Intelligence Activities`, `Church Committee Report`
- `Rockefeller Commission` — also written `Commission on CIA Activities within the United States`
- `institutional review board` — also written `ethics review board`, `human subjects protection committee`
- `front organizations` — also written `cutout organizations`, `proprietary funding conduits`
- `subproject` — also written `MKUltra subproject`, `numbered subproject`
- `Freedom of Information Act release` — also written `FOIA documents`, `declassified CIA records`
- `Cold War intelligence oversight` — also written `intelligence community oversight`, `congressional intelligence oversight`
- `psychological manipulation` — also written `covert psychological experimentation`, `mind control experimentation`
- `biomedical ethics violations` — also written `research ethics violations`, `Nuremberg Code violations`

### Institutions named as places to look

- Central Intelligence Agency
- Church Committee (U.S. Senate Select Committee on Intelligence)
- Rockefeller Commission
- U.S. Senate Select Committee on Intelligence
- U.S. House Permanent Select Committee on Intelligence
- National Institute of Mental Health
- Department of Defense (Army Chemical Corps)
- National Security Archive
- American Civil Liberties Union
- National Academy of Sciences (Institute of Medicine)
- Allan Memorial Institute, McGill University
- Georgetown University Hospital
- U.S. National Archives and Records Administration

### Venues named as places to look

- Journal of the History of Medicine and Allied Sciences
- Bulletin of the History of Medicine
- Intelligence and National Security
- Journal of Cold War Studies
- American Journal of Psychiatry
- Hastings Center Report
- Journal of Law, Medicine & Ethics
- Congressional hearing records and reports (U.S. Government Printing Office)
- National Security Archive Electronic Briefing Books
- Studies in Intelligence (CIA Center for the Study of Intelligence)

### Kinds of review hunted for

- Church Committee final report (1976) and related hearing transcripts
- Rockefeller Commission report (1975)
- Government Accountability Office or Congressional Research Service reports on CIA human experimentation
- Historical reviews of Cold War-era biomedical ethics violations
- Meta-analyses or historiographical reviews of declassified MKUltra financial and subproject documents
- Institute of Medicine or National Academies reports on human subjects protections arising from Cold War-era abuses
- Legal and bioethics literature reviews on informed consent failures in government-sponsored research
- Archival finding aids and document compilations from FOIA-released CIA records

## 6. Key Findings

Findings this run could not settle are not here; they are open questions, and
§14 holds them.

### Findings that cite evidence

- `fnd_36e0d278d5795134_00` [observed] The ledger favours H1 over both alternatives at the level it actually speaks to: unwitting administration of MKULTRA substances to U.S. citizens was a standing programme element whose continuation was formally deliberated in the office of the Deputy Director of Central Intelligence, with named senior officials (Gottlieb and Helms) arguing for continuation and others (Earman, Carter, Kirkpatrick) opposing, and with the participants agreeing the question might have to be escalated to the DCI for decision (ev_88177732ca07_01, ev_88177732ca07_02, ev_88177732ca07_05, ev_88177732ca07_06). A practice that requires DDCI- and DCI-level adjudication and a formal 'freeze' is a managed policy, not an accident of oversight. Cites `ev_88177732ca07_01`, `ev_88177732ca07_02`, `ev_88177732ca07_05`, `ev_88177732ca07_06`.
- `fnd_71b81992bd8ad0f3_01` [inferred] H3's premise that the absence of ethical review reflected merely undeveloped contemporaneous norms is undercut by the grounds on which restraint was actually argued inside the Agency: unwitting testing on foreign nationals 'had been ruled out' because station chiefs judged it 'too dangerous' and lacked 'controlled facilities' (ev_88177732ca07_03), the domestic safehouse arrangements were criticised internally as slipshod (ev_88177732ca07_04), and continuation was framed as a 'problem' warranting director-level sign-off (ev_88177732ca07_05). The recorded calculus is operational risk and exposure, and the recorded objection is internal — which means an ethical objection existed within the institution at the time, contrary to a 'no norms yet' account. Cites `ev_88177732ca07_03`, `ev_88177732ca07_04`, `ev_88177732ca07_05`, `ev_88177732ca07_02`.
- `fnd_4b5395942e8f1a8a_03` [inferred] H2's core quantitative claim is untouched rather than refuted by this ledger, and this is the sharpest limit on the whole reading: every item is drawn from one policy-level deliberation and its 1975 cover memo (ev_88177732ca07_01, ev_88177732ca07_02, ev_88177732ca07_05, ev_88177732ca07_06). There is no subject-level record, no consent documentation, no subproject inventory and hence no denominator, so the ledger cannot say what proportion of the ~150 subprojects involved covert administration versus witting volunteers, contract university science, or animal work. Cites `ev_88177732ca07_01`, `ev_88177732ca07_02`, `ev_88177732ca07_05`, `ev_88177732ca07_06`.
- `fnd_78674fd22e1b64b5_04` [inferred] The condition under which the pro-H1 reading stops holding is specific and visible inside the ledger itself: the only concrete venue attached to 'unwitting tests' here is the domestic safehouse operation (ev_88177732ca07_04), and the foreign-national extension was rejected for want of controlled facilities (ev_88177732ca07_03). If 'unwitting testing' in this deliberation denotes that one operational line rather than a practice distributed across subprojects, then a high-level policy fight over it is fully compatible with H2's structural claim that non-consensual dosing was concentrated in a few operationally-run efforts while most subprojects were something else. Cites `ev_88177732ca07_03`, `ev_88177732ca07_04`.
- `fnd_2c345ea699008c10_05` [observed] The inspector general's review functioned as an intervention that changed practice: the IG recommendations produced the disputed meeting, the meeting produced an agreement that continuation would require escalation, and the DCI then deferred the decision for a year while directing that the Agency 'continue the freeze on unwitting testing' (ev_88177732ca07_01, ev_88177732ca07_05, ev_88177732ca07_06). The sequence is fixed in the documents (review precedes deliberation precedes freeze), and the freeze is a deliberate withholding of the practice against the advocacy of Gottlieb and Helms, i.e. an internally imposed control on the exposure rather than a drift in it. Cites `ev_88177732ca07_01`, `ev_88177732ca07_05`, `ev_88177732ca07_06`, `ev_88177732ca07_02`. Causal claim, on: temporality, controls.
- `fnd_5d4895f42c4730df_06` [inferred] The one item bearing on what happened after the freeze cannot carry weight, and the reason is a conflict internal to the ledger: the fragment reads that the 1975 memo's authors found 'no record … that this freeze was ever lifted' (ev_88177732ca07_07), whereas its recorded bearing treats the same passage as concerning a records-destruction freeze, and its bearings are scored neutral on all three hypotheses. On either reading, an absence of a record of lifting is not evidence that unwitting testing ceased, particularly in a programme whose files were largely destroyed; the deferral in ev_88177732ca07_06 was explicitly for one year only. Cites `ev_88177732ca07_07`, `ev_88177732ca07_06`.

### Hypothesized

A hypothesized finding is a candidate reading this run wrote down. The run
**did not establish it**. It may still name passages — the contract requires
evidence of the two statuses above and forbids it to the one below, and asks
nothing of this one — so a citation here shows what suggested the reading, not
what settled it. The group is kept apart for that reason: listed beside an
observed finding, a candidate reads as a conclusion the run reached.

- `fnd_5c11b3861b8be7bc_07` [hypothesized] H1 and H2 are posed as exclusive but need not be: a position in which unwitting dosing was systemic as authorised policy — deliberated, defended and frozen at DDCI/DCI level — while remaining operationally concentrated in a small number of subprojects would satisfy H1's account of oversight and consent and H2's account of the documentary distribution simultaneously. Discriminating between them requires subproject-level counts, not more policy correspondence, which is why additional documents of the kind in this ledger would not move the question.

## 7. Evidence

Excerpts are printed verbatim, inside a fence: this document writes the
stored bytes and does not edit them. Two differences belong to the format
rather than to the passage, and both show up only when a passage is read
back out of this page. A fenced block terminates each of its own lines, so
an excerpt that did not end in a newline gains one. And a markdown parser
normalises as it reads: a carriage return comes back as a newline, and a
null byte comes back as U+FFFD, because the format requires both. A digest
recomputed over text copied out of this page can therefore differ from the
snapshot digest printed beside it for those reasons alone — that digest is
the one taken over the stored bytes, and it remains the thing to check a
passage against.

### Passages weighed against every live explanation

#### `ev_88177732ca07_01`

- Answers subquestion `SQ4`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 1`

```
This memo records a meeting held in the office of Deputy Director of Central Intelligence Gen. Marshall Carter to settle the one major point of disagreement among CIA officials over the inspector general’s MKULTRA recommendations: whether to continue with the testing of MKULTRA substances on unwitting U.S. citizens.
```

#### `ev_88177732ca07_02`

- Answers subquestion `SQ4`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 2`

```
Both Gottlieb and Helms “argued for the continuation of unwitting testing,” while Earman, Carter and Kirkpatrick disagreed.
```

#### `ev_88177732ca07_03`

- Answers subquestion `SQ13`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 2`

```
Carter was concerned with the “unwitting aspect,” and a discussion ensued “on the possibility of unwitting test on foreign nationals,” which “had been ruled out” due to opposition from “senior chiefs of stations” as “too dangerous” and who said they lacked “controlled facilities.”
```

#### `ev_88177732ca07_04`

- Answers subquestion `SQ8`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 2`

```
Earman finds this “odd,” emphasizing the slipshod nature of some of the safehouses used for unwitting tests in the U.S.
```

#### `ev_88177732ca07_05`

- Answers subquestion `SQ8`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 3`

```
Concluding the meeting, the participants agree that if the Directorate for Plans determined “that unwitting testing on American citizens must be continued to operationally prove out these drugs, it may become necessary to place this problem before the Director [of Central Intelligence] for a decision.”
```

#### `ev_88177732ca07_06`

- Answers subquestion `SQ15`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 3`

```
The attached cover memo from 1975 indicates that the DCI decided to defer a decision on testing U.S. citizens for one year and requested that until then the Agency “please continue the freeze on unwitting testing.”
```

#### `ev_88177732ca07_07`

- Answers subquestion `SQ14`.
- Read from `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive
- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Located at: `Description, para. 3`

```
The authors of the cover memo found “no record … that this freeze was ever lifted.”
```

### Passages nothing has scored

These passed the verbatim gate, and nothing has compared them to the live
explanations. No finding in this document rests on one.

*This run holds no unscored passages.*

## 8. Competing Explanations

The roster in its own order, with the five bearing counts each explanation
collected. The counts are unweighted by evidence tier on purpose: no weight
vector has been argued anywhere in this system, and an invented one would let
a genre label quietly outvote the passages themselves. No condition below
carries a probability either — a narrated number moves the argument onto the
number and away from the condition nobody can support.

| Explanation | ++ | + | 0 | - | -- |
|---|---|---|---|---|---|
| `H1` | 2 | 4 | 1 | 0 | 0 |
| `H2` | 0 | 0 | 1 | 6 | 0 |
| `H3` | 0 | 0 | 1 | 4 | 2 |

### `H1` — MKUltra involved substantial, well-documented non-consensual experimentation on unwitting subjects (including covert LSD dosing of non-volunteers, prisoners, and mental patients) alongside witting research, with oversight mechanisms deliberately bypassed and no meaningful ethical review or informed consent process in place across most subprojects.

- Falsified by: Discovery that surviving Church Committee/Rockefeller Commission records and the ~20,000 recovered financial documents show that the documented cases of non-consensual dosing (e.g., Operation Midnight Climax, Frank Olson case) were isolated exceptions rather than a pattern, and that the majority of the ~150 subprojects had documented consent procedures or witting participant pools, with ethical review requirements formally in place and generally followed.
- Disconfirming evidence would be observable at: Declassified CIA MKUltra financial and subproject documents held at the National Archives; 1977 Church Committee joint hearing record ('Human Drug Testing by the CIA'); Rockefeller Commission report text
- Required conditions — **all of these must hold**, and the weakest is where
  the argument actually breaks:
  1. Church Committee/Rockefeller Commission testimony and surviving documents identify multiple specific subprojects involving unwitting subjects, not merely one or two anecdotal cases
  2. Surviving records show absence or circumvention of informed consent procedures across a substantial share of subprojects, not just in cases where records happen to survive
  3. Institutional oversight (CIA internal review, IG audits, external ethical review) is documented as either absent or actively circumvented, not merely undocumented due to record destruction
- Passages declaring a bearing on it: 7.

### `H2` — The extent of non-consensual experimentation is substantially overstated relative to documented evidence; most surviving MKUltra records describe legitimate (if ethically questionable by modern standards) research on witting volunteers, university contract science, and animal/limited human testing, with the 'unwitting subjects' narrative resting on a small number of well-publicized cases (e.g., Midnight Climax, Olson) that have been generalized to the whole 150-plus subproject program without documentary support for most subprojects.

- Falsified by: Finding that a significant proportion (not just a handful) of the documented subprojects for which subject-level records survive explicitly describe covert administration of drugs or psychological manipulation to subjects who had not consented, indicating the pattern was systemic rather than exceptional.
- Disconfirming evidence would be observable at: Church Committee victim/witness testimony transcripts; subproject-level CIA financial documents; historical scholarship (e.g., John Marks' 'The Search for the Manchurian Candidate') cataloguing subproject subject matter
- Required conditions — **all of these must hold**, and the weakest is where
  the argument actually breaks:
  1. The bulk of the ~150 subprojects for which financial/administrative records survive lack subject-level detail sufficient to characterize them as non-consensual or consensual
  2. The publicized non-consensual cases (Midnight Climax, Olson death, Canadian Allan Memorial Institute experiments) represent a small numerical fraction of total documented subprojects
  3. No independent corroborating testimony (e.g., victim accounts documented in Church Committee hearings) extends the non-consensual pattern beyond these known cases
- Passages declaring a bearing on it: 7.

### `H3` — Institutional oversight failures were not primarily due to deliberate circumvention by CIA officials but reflected the prevailing legal and regulatory environment of the 1950s-60s, in which formal human-subjects protections (IRBs, codified informed consent requirements) did not yet exist in U.S. law or practice generally, making MKUltra's lack of ethical review consistent with (not exceptional relative to) contemporaneous biomedical research norms outside the program.

- Falsified by: Documentation showing that CIA officials or contracted researchers were aware of and explicitly instructed to violate existing applicable standards (e.g., the Nuremberg Code, which the U.S. had endorsed post-WWII, or internal military/medical ethics guidance already in use elsewhere) specifically to avoid detection, indicating conscious violation rather than mere absence of yet-to-be-developed norms.
- Disconfirming evidence would be observable at: Church Committee final report sections on legal/ethical context; historical scholarship on U.S. human-subjects research regulation timeline (e.g., pre-Belmont Report era); internal CIA memoranda referencing legal risk or Nuremberg standards
- Required conditions — **all of these must hold**, and the weakest is where
  the argument actually breaks:
  1. Contemporaneous U.S. biomedical/psychological research generally lacked formalized IRB-style ethical review comparable to MKUltra's own practices during 1953-1973
  2. No internal CIA memos or Church Committee findings show explicit awareness that MKUltra methods violated known standards (e.g., Nuremberg Code) at the time they were conducted
  3. Church Committee conclusions characterize the oversight gap as systemic/era-typical rather than as deliberate concealment specific to MKUltra
- Passages declaring a bearing on it: 7.

### Explanations a count cannot settle

An explanation whose disconfirming evidence has nowhere it could be seen can
never lose, and therefore can never win: every empty search is absorbed as
further concealment. Its counts are printed above because the evidence really
was weighed against it and hiding the column would conceal that — but a count
**cannot settle** an explanation nothing could disconfirm, and the shape of
the claim is the finding here, not the number.

*This run holds no self-sealing explanations.*

## 9. Contradictory Evidence

### Adversarial searches

#### `d8fcced75c2c93187b7bc8dc` against `H1`

- Searched: MKUltra subprojects informed consent documentation proportion witting volunteer research surviving records
- Looked for: Evidence that the majority of the ~149-150 MKUltra subprojects involved documented consent, formal institutional review, or witting volunteer subjects, showing non-consensual dosing was rare.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `247062073d350e9a92dc38c5` against `H1`

- Searched: Church Committee MKUltra findings institutional review ethical compliance proportion subprojects
- Looked for: Findings in the Church Committee report or Rockefeller Commission detailing standard ethical oversight and compliance protocols across MKUltra subprojects.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `68c197182fa577442d457dc8` against `H2`

- Searched: MKUltra subprojects list unwitting human subjects covert dosing frequency systemic evidence
- Looked for: Detailed inventories or analyses of surviving MKUltra subproject files showing widespread, systemic unwitting administration across dozens of subprojects beyond Midnight Climax and Olson.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `eba505851a4823a4ac62cf3c` against `H2`

- Searched: MKUltra surviving financial records 1977 FOIA unwitting testing subprojects count
- Looked for: Analysis of the ~20,000 financial records discovered in 1977 establishing covert, non-consensual drug testing in a substantial number of distinct subprojects.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `4a6d11bf9d5adad8b11fd5d6` against `H2`

- Searched: Ewen Cameron Allan Memorial Institute MKUltra Subproject 68 patient consent status
- Looked for: Specific evidence of widespread non-consensual psychic driving and drug testing on psychiatric patients without consent in Subproject 68.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `3116abdbb21030b1d0fd9591` against `H3`

- Searched: CIA MKUltra Nuremberg Code awareness internal memo deliberate circumvention medical ethics
- Looked for: Contemporary 1950s CIA memos explicitly acknowledging the Nuremberg Code or military medical ethics regulations and directing researchers to bypass them covertly.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `ef3315870a488b4aec474667` against `H3`

- Searched: 1953 Wilson memo human experimentation informed consent CIA MKUltra compliance
- Looked for: Evidence that the Secretary of Defense (Wilson memo of 1953) established strict written consent rules in the US government and that CIA intentionally evaded this policy.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

#### `85b3b178bd866454bca80074` against `H3`

- Searched: MKUltra 1963 Inspector General report ethical standards violation legal exposure secrecy
- Looked for: Statements in the 1963 IG report indicating CIA leadership knew unwitting testing was a direct, active violation of existing legal and ethical rules rather than just a public relations risk.
- Result: returned nothing; that is a result, not a blank. The search ran and came back empty.

### Passages bearing against an explanation

| Passage | Explanation | Bearing | Why |
|---|---|---|---|
| `ev_88177732ca07_01` | `H2` | - | Shows unwitting testing was a central, deliberated policy matter at the highest levels, not a minor isolated exception. |
| `ev_88177732ca07_01` | `H3` | -- | Deliberate high-level debate over whether to continue unwitting testing shows conscious institutional choice, not mere absence of contemporaneous norms. |
| `ev_88177732ca07_02` | `H2` | - | Debate among senior officials about continuing unwitting testing suggests it was a recognized, ongoing program element, not a mere generalization from a few cases. |
| `ev_88177732ca07_02` | `H3` | - | Internal disagreement over continuing unwitting testing shows officials were aware of ethical stakes and chose to advocate for it, undercutting the 'norms hadn't developed' defense. |
| `ev_88177732ca07_03` | `H2` | - | Discussion of extending unwitting testing abroad undercuts the claim that non-consensual experimentation was isolated to a few well-known cases. |
| `ev_88177732ca07_03` | `H3` | - | The rejection was based on operational danger and lack of controlled facilities, not on legal or ethical norms, indicating conscious risk calculus rather than mere absence of ethics standards. |
| `ev_88177732ca07_04` | `H2` | - | Internal IG criticism of unwitting testing practices undercuts the claim that non-consent was an isolated, overstated phenomenon. |
| `ev_88177732ca07_04` | `H3` | - | An internal IG report criticizing procedures implies awareness of deficiencies, not mere absence of era-appropriate norms. |
| `ev_88177732ca07_05` | `H2` | - | Indicates unwitting testing was a deliberate ongoing operational policy, not an isolated exception limited to a few publicized cases. |
| `ev_88177732ca07_05` | `H3` | -- | Shows awareness that unwitting testing was a 'problem' requiring director-level sign-off, evidencing conscious internal recognition of impropriety rather than mere absence of norms. |
| `ev_88177732ca07_06` | `H2` | - | A 1975 internal memo about freezing 'unwitting testing' implies the practice was significant enough to warrant high-level policy attention, undercutting the 'isolated cases' narrative. |
| `ev_88177732ca07_06` | `H3` | - | The DCI's deliberate decision to maintain a freeze on unwitting testing shows conscious awareness and management of the practice rather than mere absence of contemporaneous ethical norms. |

### Pre-registered probes

*This run does not record any pre-registered probe.*

A probe states, before a search runs, what should be findable and where.
Only a search like that turns an empty result into evidence of absence;
without one, nothing distinguishes a thing that is not there from a place
nobody looked. No probe was pre-registered in this run, so no absence
reported anywhere in this document counts as evidence of absence.

## 10. Analysis

This section is a matrix and a causal ledger, **not a narrative**. Nothing in
this run's state performs prose analysis, and this renderer does not invent
one; what follows is what was recorded, read across rows and down columns.

### Bearing matrix

| Passage | `H1` | `H2` | `H3` |
|---|---|---|---|
| `ev_88177732ca07_01` | ++ | - | -- |
| `ev_88177732ca07_02` | + | - | - |
| `ev_88177732ca07_03` | + | - | - |
| `ev_88177732ca07_04` | + | - | - |
| `ev_88177732ca07_05` | ++ | - | -- |
| `ev_88177732ca07_06` | + | - | - |
| `ev_88177732ca07_07` | 0 | 0 | 0 |

### Why each cell reads as it does

| Passage | Explanation | Bearing | Why |
|---|---|---|---|
| `ev_88177732ca07_01` | `H1` | ++ | Confirms top CIA officials deliberated continuing unwitting testing on US citizens, showing deliberate policy choice over ethics, not accident. |
| `ev_88177732ca07_01` | `H2` | - | Shows unwitting testing was a central, deliberated policy matter at the highest levels, not a minor isolated exception. |
| `ev_88177732ca07_01` | `H3` | -- | Deliberate high-level debate over whether to continue unwitting testing shows conscious institutional choice, not mere absence of contemporaneous norms. |
| `ev_88177732ca07_02` | `H1` | + | Shows top CIA officials actively pushed to continue unwitting testing, indicating deliberate, sustained policy rather than isolated incident.imo |
| `ev_88177732ca07_02` | `H2` | - | Debate among senior officials about continuing unwitting testing suggests it was a recognized, ongoing program element, not a mere generalization from a few cases. |
| `ev_88177732ca07_02` | `H3` | - | Internal disagreement over continuing unwitting testing shows officials were aware of ethical stakes and chose to advocate for it, undercutting the 'norms hadn't developed' defense. |
| `ev_88177732ca07_03` | `H1` | + | Shows CIA leadership actively discussed unwitting testing on foreign nationals, evidencing deliberate consideration of non-consensual methods rather than accidental oversight gaps. |
| `ev_88177732ca07_03` | `H2` | - | Discussion of extending unwitting testing abroad undercuts the claim that non-consensual experimentation was isolated to a few well-known cases. |
| `ev_88177732ca07_03` | `H3` | - | The rejection was based on operational danger and lack of controlled facilities, not on legal or ethical norms, indicating conscious risk calculus rather than mere absence of ethics standards. |
| `ev_88177732ca07_04` | `H1` | + | Inspector General's own finding of odd, slipshod safehouse practices for unwitting tests supports lack of meaningful oversight. |
| `ev_88177732ca07_04` | `H2` | - | Internal IG criticism of unwitting testing practices undercuts the claim that non-consent was an isolated, overstated phenomenon. |
| `ev_88177732ca07_04` | `H3` | - | An internal IG report criticizing procedures implies awareness of deficiencies, not mere absence of era-appropriate norms. |
| `ev_88177732ca07_05` | `H1` | ++ | Shows officials explicitly planned to continue unwitting testing on Americans, confirming deliberate bypass of consent with only internal escalation, not external ethical review.² |
| `ev_88177732ca07_05` | `H2` | - | Indicates unwitting testing was a deliberate ongoing operational policy, not an isolated exception limited to a few publicized cases. |
| `ev_88177732ca07_05` | `H3` | -- | Shows awareness that unwitting testing was a 'problem' requiring director-level sign-off, evidencing conscious internal recognition of impropriety rather than mere absence of norms. |
| `ev_88177732ca07_06` | `H1` | + | Shows CIA leadership explicitly acknowledged ongoing unwitting testing and treated it as a policy matter requiring deliberate management, not mere oversight lapse. |
| `ev_88177732ca07_06` | `H2` | - | A 1975 internal memo about freezing 'unwitting testing' implies the practice was significant enough to warrant high-level policy attention, undercutting the 'isolated cases' narrative. |
| `ev_88177732ca07_06` | `H3` | - | The DCI's deliberate decision to maintain a freeze on unwitting testing shows conscious awareness and management of the practice rather than mere absence of contemporaneous ethical norms. |
| `ev_88177732ca07_07` | `H1` | 0 | A note about a 1973 freeze on document destruction says nothing about the scope or pattern of non-consensual experimentation itself. |
| `ev_88177732ca07_07` | `H2` | 0 | This fragment about a destruction freeze does not bear on whether unwitting dosing was widespread or exceptional. |
| `ev_88177732ca07_07` | `H3` | 0 | The freeze memo concerns records retention, not whether oversight failures reflected era norms or deliberate circumvention. |

### Causal ledger

What each passage's evidence establishes about causation, from a closed
vocabulary. An empty entry is the ordinary answer: most passages report an
association, and saying so is a result rather than an omission.

- `ev_88177732ca07_01` — temporality
- `ev_88177732ca07_02` — no causal criterion named; association only
- `ev_88177732ca07_03` — no causal criterion named; association only
- `ev_88177732ca07_04` — no causal criterion named; association only
- `ev_88177732ca07_05` — controls
- `ev_88177732ca07_06` — temporality
- `ev_88177732ca07_07` — no causal criterion named; association only

And what each finding claims of it:

- `fnd_36e0d278d5795134_00` — not a causal claim
- `fnd_71b81992bd8ad0f3_01` — not a causal claim
- `fnd_73f5b585b3c8652c_02` — not a causal claim
- `fnd_4b5395942e8f1a8a_03` — not a causal claim
- `fnd_78674fd22e1b64b5_04` — not a causal claim
- `fnd_2c345ea699008c10_05` — temporality, controls
- `fnd_5d4895f42c4730df_06` — not a causal claim
- `fnd_5c11b3861b8be7bc_07` — not a causal claim
- `fnd_dff08e0b1dbe3663_00` — not a causal claim
- `fnd_18bfb90d81469896_01` — not a causal claim
- `fnd_606a2e68a6aa5223_02` — not a causal claim
- `fnd_f0ee31d3ef43a7d4_03` — not a causal claim
- `fnd_93041711e6aa544f_04` — not a causal claim

## 11. Conclusions

There is **no run-level verdict** in this run's state — no field holds one —
and this renderer does not invent one. What follows is what the run recorded,
each claim carrying the kind of claim it is.

*This run does not record which stage minted each finding.*

`ResearchState.findings` is a single tuple with no stage marker, so a reloaded
run cannot separate the synthesis from the later gaps sweep. The split between
this section and §14 therefore rests on each finding's own status, which is a
property of the claim, not on which stage produced it.

### What the run recorded

- `fnd_36e0d278d5795134_00` [observed] The ledger favours H1 over both alternatives at the level it actually speaks to: unwitting administration of MKULTRA substances to U.S. citizens was a standing programme element whose continuation was formally deliberated in the office of the Deputy Director of Central Intelligence, with named senior officials (Gottlieb and Helms) arguing for continuation and others (Earman, Carter, Kirkpatrick) opposing, and with the participants agreeing the question might have to be escalated to the DCI for decision (ev_88177732ca07_01, ev_88177732ca07_02, ev_88177732ca07_05, ev_88177732ca07_06). A practice that requires DDCI- and DCI-level adjudication and a formal 'freeze' is a managed policy, not an accident of oversight. Cites `ev_88177732ca07_01`, `ev_88177732ca07_02`, `ev_88177732ca07_05`, `ev_88177732ca07_06`.
- `fnd_71b81992bd8ad0f3_01` [inferred] H3's premise that the absence of ethical review reflected merely undeveloped contemporaneous norms is undercut by the grounds on which restraint was actually argued inside the Agency: unwitting testing on foreign nationals 'had been ruled out' because station chiefs judged it 'too dangerous' and lacked 'controlled facilities' (ev_88177732ca07_03), the domestic safehouse arrangements were criticised internally as slipshod (ev_88177732ca07_04), and continuation was framed as a 'problem' warranting director-level sign-off (ev_88177732ca07_05). The recorded calculus is operational risk and exposure, and the recorded objection is internal — which means an ethical objection existed within the institution at the time, contrary to a 'no norms yet' account. Cites `ev_88177732ca07_03`, `ev_88177732ca07_04`, `ev_88177732ca07_05`, `ev_88177732ca07_02`.
- `fnd_4b5395942e8f1a8a_03` [inferred] H2's core quantitative claim is untouched rather than refuted by this ledger, and this is the sharpest limit on the whole reading: every item is drawn from one policy-level deliberation and its 1975 cover memo (ev_88177732ca07_01, ev_88177732ca07_02, ev_88177732ca07_05, ev_88177732ca07_06). There is no subject-level record, no consent documentation, no subproject inventory and hence no denominator, so the ledger cannot say what proportion of the ~150 subprojects involved covert administration versus witting volunteers, contract university science, or animal work. Cites `ev_88177732ca07_01`, `ev_88177732ca07_02`, `ev_88177732ca07_05`, `ev_88177732ca07_06`.
- `fnd_78674fd22e1b64b5_04` [inferred] The condition under which the pro-H1 reading stops holding is specific and visible inside the ledger itself: the only concrete venue attached to 'unwitting tests' here is the domestic safehouse operation (ev_88177732ca07_04), and the foreign-national extension was rejected for want of controlled facilities (ev_88177732ca07_03). If 'unwitting testing' in this deliberation denotes that one operational line rather than a practice distributed across subprojects, then a high-level policy fight over it is fully compatible with H2's structural claim that non-consensual dosing was concentrated in a few operationally-run efforts while most subprojects were something else. Cites `ev_88177732ca07_03`, `ev_88177732ca07_04`.
- `fnd_2c345ea699008c10_05` [observed] The inspector general's review functioned as an intervention that changed practice: the IG recommendations produced the disputed meeting, the meeting produced an agreement that continuation would require escalation, and the DCI then deferred the decision for a year while directing that the Agency 'continue the freeze on unwitting testing' (ev_88177732ca07_01, ev_88177732ca07_05, ev_88177732ca07_06). The sequence is fixed in the documents (review precedes deliberation precedes freeze), and the freeze is a deliberate withholding of the practice against the advocacy of Gottlieb and Helms, i.e. an internally imposed control on the exposure rather than a drift in it. Cites `ev_88177732ca07_01`, `ev_88177732ca07_05`, `ev_88177732ca07_06`, `ev_88177732ca07_02`. Causal claim, on: temporality, controls.
- `fnd_5d4895f42c4730df_06` [inferred] The one item bearing on what happened after the freeze cannot carry weight, and the reason is a conflict internal to the ledger: the fragment reads that the 1975 memo's authors found 'no record … that this freeze was ever lifted' (ev_88177732ca07_07), whereas its recorded bearing treats the same passage as concerning a records-destruction freeze, and its bearings are scored neutral on all three hypotheses. On either reading, an absence of a record of lifting is not evidence that unwitting testing ceased, particularly in a programme whose files were largely destroyed; the deferral in ev_88177732ca07_06 was explicitly for one year only. Cites `ev_88177732ca07_07`, `ev_88177732ca07_06`.
- `fnd_5c11b3861b8be7bc_07` [hypothesized] H1 and H2 are posed as exclusive but need not be: a position in which unwitting dosing was systemic as authorised policy — deliberated, defended and frozen at DDCI/DCI level — while remaining operationally concentrated in a small number of subprojects would satisfy H1's account of oversight and consent and H2's account of the documentary distribution simultaneously. Discriminating between them requires subproject-level counts, not more policy correspondence, which is why additional documents of the kind in this ledger would not move the question.

### How the explanations stood at the end

In the roster's own order, deliberately not sorted by any summed score.

| Explanation | ++ | + | 0 | - | -- |
|---|---|---|---|---|---|
| `H1` | 2 | 4 | 1 | 0 | 0 |
| `H2` | 0 | 0 | 1 | 6 | 0 |
| `H3` | 0 | 0 | 1 | 4 | 2 |

## 12. Confidence Assessment

Every level below is **computed** from this run's own ledger: how many
passages back the finding, how many distinct *origins* those passages trace
to, and whether the counter-search came back holding anything.
**No stage was asked for this label**, and no model typed one. Origins rather
than hosts, because four hundred sites paraphrasing one blog post are one
source; and a source whose provenance could not be established never raises a
level, since unknown provenance is unknown independence.

A finding that cites nothing reads `unknown`. That is the honest result and
not a missing measurement: **nothing in this run backs it** yet.

| Finding | Level | What that is counted from |
|---|---|---|
| `fnd_36e0d278d5795134_00` | unknown | 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin. |
| `fnd_71b81992bd8ad0f3_01` | unknown | 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin. |
| `fnd_73f5b585b3c8652c_02` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |
| `fnd_4b5395942e8f1a8a_03` | unknown | 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin. |
| `fnd_78674fd22e1b64b5_04` | unknown | 2 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin. |
| `fnd_2c345ea699008c10_05` | unknown | 4 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin. |
| `fnd_5d4895f42c4730df_06` | unknown | 2 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced; 1 source(s) could not be traced to an origin. |
| `fnd_5c11b3861b8be7bc_07` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |
| `fnd_dff08e0b1dbe3663_00` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |
| `fnd_18bfb90d81469896_01` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |
| `fnd_606a2e68a6aa5223_02` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |
| `fnd_f0ee31d3ef43a7d4_03` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |
| `fnd_93041711e6aa544f_04` | unknown | 0 passage(s) tracing to 0 distinct origin(s); no counter-evidence surfaced. |

## 13. Limitations

### Audit measurements

Counted from the ledger rather than answered by a model, and printed even
when the result is zero. Provenance tracing and primary-data classification
are separate measurements: reaching an upstream origin does not make a source
primary data, and a primary-data source can still have unresolved provenance.

- Origin tracing: 0/4 snapshot(s) reached an upstream origin (0.00%).
- Primary-data share: 0/4 assessed source(s) were assigned stage 9's `primary_data` tier (0.00%).

#### Distinct source families behind each asserted finding

Repeated passages and several URLs from one family still count once. Unknown
findings are open questions rather than conclusions, so they are not rows here.

| Finding | Distinct source families |
|---|---|
| `fnd_36e0d278d5795134_00` | 1 |
| `fnd_71b81992bd8ad0f3_01` | 1 |
| `fnd_4b5395942e8f1a8a_03` | 1 |
| `fnd_78674fd22e1b64b5_04` | 1 |
| `fnd_2c345ea699008c10_05` | 1 |
| `fnd_5d4895f42c4730df_06` | 1 |
| `fnd_5c11b3861b8be7bc_07` | 0 |

- Falsification yield: 0/8 adversarial search(es) yielded a snapshot-bound, verified counter-passage (0.00%).

- Excerpt round-trip: 7/7 readable passage(s) survived the verbatim gate (100.00%); 0 were refused and 0 could not be checked.

### Attacks made on this run's own conclusions

- `stp_2e9d970e515e43db_00` [causal_overreach] against `fnd_36e0d278d5795134_00`: High-level DDCI deliberation and an executive freeze do not establish that unwitting administration was an established, managed policy program-wide; executive intervention may have been an ad hoc containment response to rogue or compartmentalized activities brought to light by the IG. It fails if: Contemporaneous records demonstrate that the DDCI meeting was convened strictly as crisis containment for isolated, unauthorized safehouse experiments rather than a managed review of an authorized Agency-wide policy.
- `stp_a7fee67587392a95_01` [alternative_explanation] against `fnd_71b81992bd8ad0f3_01`: Internal objections grounded in operational danger, lack of controlled facilities, and slipshod safehouses reflect tactical risk management and fear of public exposure, not contemporaneous biomedical ethical norms. It fails if: Archival evidence demonstrates that internal pushback was motivated exclusively by security vulnerability, diplomatic blowback, and legal exposure without reference to ethical or human-subjects protection principles.
- `stp_87495abc063099c9_02` [source_dependence] against `fnd_4b5395942e8f1a8a_03`: Concluding that H2's quantitative distribution cannot be evaluated relies on the ledger's single executive memo chain, ignoring external documentary streams such as surviving subproject financial ledgers or institutional records. It fails if: Financial disbursement records, contract files, or institutional archives provide a verifiable denominator showing what proportion of subprojects engaged in non-consensual testing versus standard academic research.
- `stp_0f0874c1612a0843_03` [selection_bias] against `fnd_78674fd22e1b64b5_04`: Limiting the scope of unwitting tests to domestic safehouses assumes the deliberation records surveyed all venues, whereas non-consensual testing may have been dispersed across academic, medical, and penal subprojects outside safehouse channels. It fails if: Primary subproject records establish that non-consensual drug administration occurred across diverse institutional settings beyond the domestic safehouses.
- `stp_0043f38c92885011_04` [causal_overreach] against `fnd_2c345ea699008c10_05`: The finding assumes the formal DCI freeze successfully altered field practice without verifying compliance or operational adherence at the safehouse level. It fails if: Field logs or operational records indicate that covert administration of substances continued at safehouse facilities during the nominal freeze period.
- `stp_c9f025c35b65201a_05` [measurement_error] against `fnd_5d4895f42c4730df_06`: Dismissing the probative value of the 1975 memo's remark about the unlifted freeze rests on perceived ambiguity in the ledger's encoding rather than an actual substantive contradiction in the underlying document. It fails if: Unredacted archival context confirms that the 1975 memo explicitly referenced the testing freeze rather than a records freeze and verified that unwitting administration was never formally reauthorized.
- `stp_7f3bae616a7bef74_06` [alternative_explanation] against `fnd_5c11b3861b8be7bc_07`: Claiming that discriminating between H1 and H2 strictly requires subproject-level counts overlooks that program-level charters or policy directives could establish programmatic intent and systemic policy regardless of the numeric proportion of subprojects. It fails if: Foundational program charters or overarching authorization directives reveal that non-consensual administration was codified as a systemic capability requirement across MKULTRA rather than an isolated subproject anomaly.

### Saturation

Measured, never judged. This pipeline retrieves exactly twice, so the only
comparison available is round 1 alone against both rounds combined. Source
exhaustion and information exhaustion are different things, and this measures
the first.

- Round 1 found 3 source family(ies).
- Round 2 found 0, of which 0 were new.
- New-family share of round 2: 0.00.
- Explanations whose counts moved when round 2 was added: none.
- Saturated by idea.md's rule: yes.

### What the sources do not state

- `src_36f52cbfeddd0179482b789f` does not state: The document does not state who compiled the original 1976 briefing book beyond CIA, does not state the identity or credentials of whoever transcribed/OCR'd it for Internet Archive, does not state verification procedures for transcription accuracy, and does not state a peer-review or editorial process.
  - Evidence-to-conclusion gap: The source is a raw archival transcription with no stated conclusions of its own; it merely reproduces primary CIA administrative records without analysis, so there is no gap between evidence and conclusion to observe.
- `src_378c09c6004245fc5796c9ea` does not state: The report does not state peer review, replication, or independent verification of its findings by outside parties, nor inter-rater reliability of its document analysis.
  - Evidence-to-conclusion gap: The report's sweeping conclusion that constitutional rights were 'threatened and undermined' rests on documented specific incidents; some dissenting Senators argued the recommendations extended well beyond what the assembled record actually supported.
- `src_7b19358adeeae5762915db89` does not state: Does not state outcomes of unwitting tests, later decisions after the one-year deferral, or verification of whether the freeze was maintained beyond noting no record was found lifting it.
  - Evidence-to-conclusion gap: The memo's own conclusions are narrow (a meeting record); the National Security Archive's framing draws broader inferences about CIA policy and testing practices that extend beyond what this single document directly documents.
- `src_d00b850ba8c967e4b32cf684` does not state: Sampling methodology and selection criteria for SIS and RPRP, statistical analysis methods, funding amounts, conflict-of-interest disclosures, and criteria for selecting which archival case studies to feature are not stated.
  - Evidence-to-conclusion gap: The report's recommendations extend broadly to future federal policy and calls for cultural change in the research community, which go beyond the historical case studies and two contemporary projects described as the evidentiary basis; this is a broader normative and policy conclusion than the descriptive/documentary evidence presented.

### Provenance not established

- 4 source(s) could not be traced to an origin.

*This run does not record how each source's provenance was resolved.*

Only the resulting origin id reaches canonical state, so this document
**cannot tell an original source from an untraced one**: both arrive holding
no origin, and unknown provenance is unknown independence.

### Explanations no observation could settle

*This run holds no self-sealing explanations.*

### What the question assumed

- This scope treats 'documented' evidence as that surfacing from the 1977 Senate Church Committee hearings, the 1975 Rockefeley Commission, and the roughly 20,000 surviving financial documents located after the 1973 destruction order, rather than speculative or unverified claims
- Related precursor and successor programs (BLUEBIRD, ARTICHOKE) are treated as contextual background rather than within the core population unless explicitly designated as MKUltra subprojects

### What this document cannot say

*This run does not record any pre-registered probe.*

*This run does not record whether any reviewer approved any gate.*

*This run does not record its own cost or wall-clock time.*

The gate record and the budget counters live in the console and are not
carried by canonical state, so nothing in this document is a claim that a
person reviewed any step of this run.

One more, about this page rather than about the run. No URL written by a
model is a link here: a source's address is printed as text beside the
digest of the bytes that were actually stored, because a clickable target
is a claim that the address resolves to what the run read, and nothing
checked that. An email address typed into a recorded fact is the single
exception a markdown reader may still turn into a link on its own, and it
is not one this renderer can suppress without altering the text it was
given.

## 14. Open Questions

*This run does not record which stage minted each finding.*

As in §11: the split between a conclusion and a gap rests on each finding's
own status, because state carries no marker for which stage minted it.

### Findings this run could not settle

- `fnd_73f5b585b3c8652c_02` H3 is weakened but not falsified on this ledger: nothing here shows CIA officials or contractors being made aware of, and instructed to violate, an applicable external standard (the Nuremberg Code, service medical-ethics guidance) specifically to avoid detection, which is H3's own stated falsifier. Absent such a document, the distinction between 'conscious violation of a recognised standard' and 'conscious management of an operational and reputational hazard in a normless environment' remains open. Would be settled by: Contemporaneous CIA or contractor correspondence, legal-office opinions, or subproject proposals that cite the Nuremberg Code, service medical-ethics guidance, or an internal consent requirement, and show it being set aside or concealed.
- `fnd_dff08e0b1dbe3663_00` Whether unwitting testing resumed, continued informally, or actually ceased after the one-year freeze deferred by the DCI in 1975 remains unresolved. Would be settled by: CIA correspondence, IG reports, or Church/Rockefeller Commission records from after 1976 documenting whether the freeze was extended, lifted, or the underlying subprojects terminated.
- `fnd_18bfb90d81469896_01` The proportion of the roughly 150 MKUltra subprojects that involved covert, non-consensual administration of substances versus witting volunteers or contract research is not established by anything in this ledger. Would be settled by: A subproject-level inventory or the surviving ~20,000 financial documents cross-referenced with subject consent status for each subproject.
- `fnd_606a2e68a6aa5223_02` Whether CIA officials or contracted researchers were specifically aware of and instructed to violate an existing external ethical standard (e.g., the Nuremberg Code or military/medical ethics guidance) in order to avoid detection is unresolved. Would be settled by: Internal directives, training materials, or legal counsel memos referencing the Nuremberg Code or comparable standards in connection with MKUltra subproject design or concealment.
- `fnd_f0ee31d3ef43a7d4_03` Whether the 'unwitting testing' discussed in the 1963 IG deliberation refers narrowly to the domestic safehouse operation (e.g., Operation Midnight Climax) or to a broader set of subprojects across the program is not settled. Would be settled by: Subproject identifiers or operational descriptions linking the deliberation's subject matter to specific named subprojects beyond the safehouse operation.
- `fnd_93041711e6aa544f_04` Whether formal ethical review or informed consent procedures were documented and generally followed in the majority of subprojects, as H1's falsifier and H2 both require to be tested, is not addressed by any evidence here. Would be settled by: Surviving subproject files, contracts, or consent forms showing presence or absence of ethical review procedures across the subproject population.

### Explanations nothing has weighed

A live explanation no passage declared a bearing on. This is not the same as
an explanation every passage found non-discriminating, which is a finding.

*This run holds no explanations nothing has weighed.*

### Parts of the question no passage answers

- `SQ1` — What were the officially stated objectives and goals of Project MKUltra as documented in surviving CIA records and Congressional testimony?
- `SQ2` — What specific research methods and techniques (e.g., drug administration, sensory deprivation, hypnosis, electroshock) were used across documented MKUltra subprojects?
- `SQ3` — How many subprojects and institutions (universities, hospitals, prisons, front organizations) are documented as receiving CIA funding under MKUltra, based on surviving financial records?
- `SQ5` — What specific cases or subprojects are documented as involving covert administration of drugs (e.g., LSD) to unwitting subjects, and what were the reported consequences for those subjects?
- `SQ6` — What formal ethical review, legal authorization, or informed consent procedures (if any) governed MKUltra research activities at the time they were conducted?
- `SQ7` — In what ways did MKUltra activities circumvent or violate existing legal and ethical standards (e.g., the Nuremberg Code) applicable to human subjects research during this period?
- `SQ9` — How did the 1973 destruction of records affect the ability of subsequent investigators (Rockefeller Commission, Church Committee) to document the full scope of MKUltra's human subjects research?
- `SQ10` — What findings and conclusions did the Rockefeller Commission (1975) and Church Committee (1977) reach regarding the extent of non-consensual experimentation and oversight failures in MKUltra?
- `SQ11` — Where do sources (Church Committee, Rockefeller Commission, FOIA-released documents) disagree on the scope, number of subjects, or extent of non-consensual experimentation under MKUltra?
- `SQ12` — What role did precursor programs (BLUEBIRD, ARTICHOKE) play in establishing methods or objectives later continued under MKUltra, insofar as they are documented as directly linked to MKUltra subprojects?

### Ambiguities the question recorded

- The full scope of MKUltra remains uncertain because most records were ordered destroyed in 1973; conclusions may differ depending on whether the question is scoped to documented/surviving records only or presumed full extent
- Whether 'MKUltra' should be treated as coextensive with the broader family of related CIA programs (e.g., MKDELTA, MKNAOMI, Project ARTICHOKE) or as a narrower, distinct program affects population and period boundaries
- The number of subprojects and total human subjects affected varies across sources (Church Committee, Rockefeller Commission, subsequent FOIA releases), which would change quantitative answers

## 15. Sources

*This run does not record which query found each source.*

Retrieval keeps no link from a search to the sources it returned, and this
renderer will not invent one; §4's searches and the list below cannot be
joined by anything in this document.

### Snapshotted sources

Bytes were fetched, hashed and stored for each of these. Only these can
support a claim.

#### `src_36f52cbfeddd0179482b789f` — Full text of " MKULTRA Briefing Book List Of Subprojects With Brief Descriptions January 1976 "

- URL: `https://archive.org/stream/MKULTRABriefingBookListOfSubprojectsWithBriefDescriptionsJanuary1976/MKULTRA%20Briefing%20Book%20-%20List%20of%20subprojects%20with%20brief%20descriptions%20-%20January%201976_djvu.txt`
- Retrieval round: 1.
- Source family: `web:archive.org`.
- Origin: not established.
- Snapshot digest: `3e623f41b22a02f083981d8f8a4ebe1cee494f4b3493d0bda549cffed622dec1`
- Fetched at: 2026-08-29T18:33:11.611227+00:00
- Venue: not recorded.
- Authors: not recorded
- Published: not recorded
- Retracted: no.
- Assessed tier: `unsupported`.
- What was studied: The source is an Internet Archive-hosted OCR transcription of a declassified CIA MKULTRA briefing book listing 149 subprojects with brief descriptions of researchers, objectives, funding, and CIA monitors.
- Data collection: Content consists of scanned/OCR'd internal CIA memoranda, budget statements, correspondence, and project summaries compiled in 1976 as an internal briefing document, later declassified and posted online.
- Sample: Covers 149 numbered subprojects; no defined sampling methodology since this is an administrative/historical compilation of internal records rather than a study with a study population.
- Controls: None stated; this is a record-keeping document, not an experimental study with controls.
- Statistical approach: None stated; no statistical analysis is presented anywhere in the document.
- Funding: Funding figures for each subproject are given (CIA appropriated funds, often via cutout mechanisms like Geschickter Fund, Society for Investigation of Human Ecology), but no overall funding statement for the document's compilation or publication is given.
- Conflicts of interest: Not stated. The document does not disclose any conflicts of interest regarding its own compilation, transcription, or hosting; it is a reproduction of CIA-authored records with no independent editorial commentary.
- Does not state: The document does not state who compiled the original 1976 briefing book beyond CIA, does not state the identity or credentials of whoever transcribed/OCR'd it for Internet Archive, does not state verification procedures for transcription accuracy, and does not state a peer-review or editorial process.
- Evidence-to-conclusion gap: The source is a raw archival transcription with no stated conclusions of its own; it merely reproduces primary CIA administrative records without analysis, so there is no gap between evidence and conclusion to observe.

#### `src_378c09c6004245fc5796c9ea` — Full text of "Final report of the Select Committee to Study ...

- URL: `https://archive.org/stream/finalreportofsel02unit/finalreportofsel02unit_djvu.txt`
- Retrieval round: 1.
- Source family: `web:archive.org`.
- Origin: not established.
- Snapshot digest: `9b34d91504d1ffd179a9d5b6ecff9fecc938963a3957d3535c0e59ec2a674b9c`
- Fetched at: 2026-08-29T18:33:12.814827+00:00
- Venue: not recorded.
- Authors: not recorded
- Published: not recorded
- Retracted: no.
- Assessed tier: `government_institutional`.
- What was studied: A year-long congressional investigation into U.S. intelligence agencies' (FBI, CIA, NSA, military intelligence) domestic activities, examining decades of surveillance, covert action, and their legal and constitutional bases.
- Data collection: Committee reviewed roughly 110,000 pages of documents from agencies and the White House, took testimony from ~800 witnesses (250 under oath in executive session, 50 public), and compiled ~30,000 transcript pages.
- Sample: Not a sample-based study; it examined the full available documentary record and witness testimony the Committee could obtain across seven presidential administrations and multiple named agencies.
- Controls: No experimental controls apply; as an investigative inquiry it cross-checked documents against testimony and noted where agencies withheld, sanitized, or destroyed records.
- Statistical approach: None; the report is a narrative and documentary analysis with findings and recommendations rather than quantitative/statistical analysis, though it cites some statistics (e.g., file counts, budget figures) drawn from agency records.
- Funding: Congressional appropriation to the Senate Select Committee (contingent fund of the Senate), capped at $750,000 through September 1975 per the authorizing resolution.
- Conflicts of interest: Not stated; as a Senate committee report, no conflict-of-interest disclosures are given, though several individual Senators appended separate/dissenting views reflecting differing institutional and political perspectives.
- Does not state: The report does not state peer review, replication, or independent verification of its findings by outside parties, nor inter-rater reliability of its document analysis.
- Evidence-to-conclusion gap: The report's sweeping conclusion that constitutional rights were 'threatened and undermined' rests on documented specific incidents; some dissenting Senators argued the recommendations extended well beyond what the assembled record actually supported.

#### `src_7b19358adeeae5762915db89` — John S. Earman, Inspector General, U.S. Central Intelligence Agency, Memorandum for the Record, “MKULTRA Program,” Secret, November 29, 1963, 2 pp. | National Security Archive

- URL: `https://nsarchive.gwu.edu/document/32732-document-17-john-s-earman-inspector-general-us-central-intelligence-agency`
- Retrieval round: 1.
- Source family: `web:nsarchive.gwu.edu`.
- Origin: not established.
- Snapshot digest: `88177732ca07983f586e559454d5a1e70cf2f6f073e361bf42872274b713d129`
- Fetched at: 2026-08-29T18:33:13.077554+00:00
- Venue: not recorded.
- Authors: not recorded
- Published: not recorded
- Retracted: no.
- Assessed tier: `government_institutional`.
- What was studied: Internal CIA record of a 1963 meeting among senior officials debating whether to continue unwitting drug testing on U.S. citizens under MKULTRA.
- Data collection: A contemporaneous memorandum for the record written by the CIA Inspector General summarizing statements and positions of named officials at a specific meeting.
- Sample: Not a study; the 'sample' is the set of participants at one meeting (Carter, Helms, Kirkpatrick, Earman, Gottlieb), whose views are recorded.
- Controls: None stated; this is an internal deliberative memo, not an experiment or analysis with controls.
- Statistical approach: None stated; no statistical methods used or applicable.
- Funding: Not stated; document is an internal CIA administrative record, not describing external funding.
- Conflicts of interest: Not stated explicitly, though the memo itself shows Gottlieb and Helms as interested parties advocating continuation of the program they ran.
- Does not state: Does not state outcomes of unwitting tests, later decisions after the one-year deferral, or verification of whether the freeze was maintained beyond noting no record was found lifting it.
- Evidence-to-conclusion gap: The memo's own conclusions are narrow (a meeting record); the National Security Archive's framing draws broader inferences about CIA policy and testing practices that extend beyond what this single document directly documents.

#### `src_d00b850ba8c967e4b32cf684` — Advisory Committee on Human Radiation Experiments

- URL: `https://sitearchives.georgetown.edu/kie-achre/final/index.html`
- Retrieval round: 1.
- Source family: `web:sitearchives.georgetown.edu`.
- Origin: not established.
- Snapshot digest: `05d23a249387455323ffa91898ce23793403bed8bd8e57e4afba48eb1adfc5b3`
- Fetched at: 2026-08-29T18:33:15.340070+00:00
- Venue: not recorded.
- Authors: not recorded
- Published: not recorded
- Retracted: no.
- Assessed tier: `government_institutional`.
- What was studied: Human radiation experiments and intentional environmental releases conducted or sponsored by the US government from 1944-1974, plus contemporary federal human subjects protections; includes case studies (plutonium injections, radioisotope distribution, children, TBI, prisoners, atomic veterans, intentional releases, uranium miners, Marshallese) and two original projects: a Research Proposal Review Project and a Subject Interview Study.
- Data collection: Archival records (including newly declassified documents), an Ethics Oral History Project, document review of IRB research proposal files (RPRP), and structured interviews with subjects (SIS). Methods for locating and selecting archival records are not detailed here.
- Sample: SIS involved interviews with nearly 1,900 patients nationwide (subjects and non-subjects). RPRP reviewed documents from research projects at institutions nationwide, radiation and nonradiation. Exact selection criteria, response rates, and sampling frame are not stated in this summary.
- Controls: No comparison/control groups described for the historical case studies, which are narrative/documentary reconstructions; SIS and RPRP appear descriptive rather than controlled analyses. No stated methodological controls for bias in archival interpretation.
- Statistical approach: Not stated in this document; no quantitative/statistical methods, tests, or analytic techniques are described for RPRP or SIS findings.
- Funding: Not stated; the report is produced by a Presidential advisory committee, implying federal funding, but no funding figures or mechanisms are given.
- Conflicts of interest: Not stated. No disclosure of committee members' potential conflicts is included in this summary document.
- Does not state: Sampling methodology and selection criteria for SIS and RPRP, statistical analysis methods, funding amounts, conflict-of-interest disclosures, and criteria for selecting which archival case studies to feature are not stated.
- Evidence-to-conclusion gap: The report's recommendations extend broadly to future federal policy and calls for cultural change in the research community, which go beyond the historical case studies and two contemporary projects described as the evidentiary basis; this is a broader normative and policy conclusion than the descriptive/documentary evidence presented.

### Indexed sources

Known by metadata alone — no bytes were fetched, so an indexed record
**cannot support a claim**. These inform the citation graph and nothing else,
and no finding in this document rests on one.

- `src_120f1721260dec161c5f8432` — Towards new human rights in the age of neuroscience and neurotechnology (`https://doi.org/10.1186/s40504-017-0050-1`), citations: 684
- `src_1da9ed3f1a21153028fee1a9` — Murky Projects and Uneven Information Policies: A Case Study of the Psychological Strategy Board and CIA (`https://doi.org/10.31979/2377-6188.2018.010206`), citations: 18
- `src_4a873a4325670fcef1aaabb6` — The Global Traffic in Human Organs (`https://doi.org/10.1086/300123`), citations: 718
- `src_ab4f0a4a0efeea20099ce586` — Current challenges in the physics of white dwarf stars (`https://doi.org/10.1016/j.physrep.2022.09.001`), citations: 137
- `src_e19665d80e0e39e7718b0046` — The war inside your mind: unprotected brain battlefields and neuro-vulnerability (`https://doi.org/10.20935/acadbiol6156`), citations: 6
- `src_f240ff52c71bf5b595fbedea` — Ethical Failures and History Lessons: The U.S. Public Health Service Research Studies in Tuskegee and Guatemala (`https://doi.org/10.1007/bf03391665`), citations: 177