BONDI HEARING: RUBIK'S CUBE CROSS-ANALYSIS
House Judiciary Committee Oversight | February 11, 2026 | 152 Coded Exchanges | Every Dimension × Every Dimension
0. EXECUTIVE SUMMARY & ACCOUNTABILITY THEATER INDEX
26
Personal Attacks by Bondi
The Accountability Theater Index (ATI)
Composite score: 0 = perfect accountability, 100 = pure theater. Weighted formula: (non-responsive rate × 30) + (attack rate × 25) + (deflection rate × 25) + (party differential × 20)
0 - Functional
25 - Imperfect
50 - Dysfunctional
75 - Theater
100 - Collapse
ATI Score: 78.4 / 100 ... firmly in "theater" territory. The hearing produced almost no binding commitments, zero subpoena threats, and a 64-point party differential in responsiveness. Bondi used personal attacks as a procedural weapon 26 times, converting oversight time into campaign content.
Top 10 Findings
1. Two-Track Justice in Real Time: Bondi answered 93% of Republican questions responsively vs. 29% of Democratic questions. A 64-point gap. This is not noise... it is structural.
2. Personal Attacks as Filibuster: 26 personal attacks, all directed at Democrats or the one Republican (Massie) who pressed on Epstein. Attacks consumed an estimated 15-20 minutes of the 4.5-hour hearing... time that could have been used for testimony.
3. The Garland Deflection: "Did you ask Merrick Garland that?" deployed 7 times as a deflection. Functions as a procedural tu quoque, shifting temporal frame to prior administration to avoid present accountability.
4. District Crime as Weapon: 10 instances of Bondi attacking a Democratic member's home district crime stats. Every single one followed an Epstein or ICE killing question. Pattern: uncomfortable question → "Why aren't you talking about [crime in your district]?"
5. Epstein: The Black Hole Topic: 42 exchanges touched Epstein... the dominant topic. Of these, only 9 received responsive answers (21%). Epstein questions were 4x more likely to trigger personal attacks than any other topic.
6. Yes/No Questions Are Most Effective: Yes/No questions got 69% responsive answers vs. Open questions at 60% and Demands at 33%. But this is almost entirely driven by party: GOP Yes/No questions had 97% response rate.
7. The Rapid-Fire Exploit: Rep. Gill (R-TX) asked 17 yes/no questions in under 3 minutes, all about Biden DOJ failures. Bondi answered every one with "Yes" / "Absolutely." This was the most "productive" exchange... and the most scripted.
8. Survivor Testimony Ignored: Epstein survivors demonstrated twice (hand-raises) that 0 of them had been contacted by DOJ despite all reaching out. Bondi did not respond to either demonstration. The room contained more evidence of obstruction than any document.
9. Zero Binding Commitments: Across 152 exchanges, Bondi made 0 binding commitments to Democrats. The only commitments were to Republican members (provide Jack Smith info to Jordan, work on fraud in various states). "I'll look into it" appeared 0 times as a concrete promise to a Democrat.
10. The Bipartisan Anomaly: Exactly 2 moments of genuine bipartisan engagement: (1) Bondi's empathetic response to Lieu's death threats, and (2) agreement with Garcia on AI/cyber child exploitation. Both occurred when the topic shifted from accountability to shared threat.
1. PARTY DIFFERENTIAL: THE TWO HEARINGS
Democrats Asked
41
Questions / Demands
Republicans Asked
48
Questions / Demands
The 64-Point Gap: 92% vs 29%. If this were a clinical trial, the treatment effect would be statistically significant at p < 0.0001. The witness is operating two entirely different protocols depending on who is asking. This is the structural signature of accountability theater... the hearing format creates the appearance of equal access while producing radically unequal outcomes.
Attack Distribution: Of 26 personal attacks, 24 (92%) were directed at Democratic members or their questions. The remaining 2 targeted Republican Thomas Massie... the only Republican who pressed aggressively on Epstein. The attack function is therefore not partisan per se, but anti-accountability: anyone who presses uncomfortable questions gets attacked regardless of party.
Reclaim Time as Proxy: Democrats had to reclaim their time 19 times vs. Republicans 3 times. Reclaiming time is a procedural indicator that the witness is running out the clock... filibustering the questioner's allotted time. This means Bondi was actively consuming Democratic questioning time at 6x the rate she consumed Republican time.
2. TOPIC ANALYSIS: WHAT GETS ANSWERED
Exchanges by Topic (All 152)
Response Rate by Topic (Questions/Demands Only)
Green = Responsive Rate | Red = Non-Responsive Rate
The Topic Gradient: Topics sort into three tiers by responsiveness. Tier 1 (80-100%): Energy, Crime Stats, Sanctuary, Immigration, Fraud, Church... all Republican priority topics. Tier 2 (40-67%): Weaponization, Jack Smith, DOJ Surveillance... mixed but tilted R. Tier 3 (0-25%): Epstein, Trump-Epstein, ICE Killings, Domestic Terrorism, Crypto, Emoluments... all Democratic priority topics. The correlation between topic ownership and response rate is nearly perfect.
Epstein Paradox: The most-discussed topic (42 exchanges, 28% of hearing) had the lowest response rate among high-volume topics (21%). Bondi spent more time deflecting Epstein questions than answering all energy, fraud, and church questions combined. The hearing's dominant subject was also its least productive.
3. RESPONSE ARSENAL: BONDI'S TACTICAL TOOLKIT
Response Types Used (Questions/Demands, n=89)
PENDING_LITIGATION4
4 (4%)
Response Type by Party
To Democrats (n=41 questions)
To Republicans (n=48 questions)
The Toolkit Split: Bondi's response to Democrats draws from 9 different response types, distributed roughly evenly... a chaos strategy designed to prevent questioners from developing countermeasures. Her response to Republicans draws from essentially 1 type (DIRECT_ANSWER at 81%). She never refused a Republican question, never attacked a Republican questioner (except Massie), and never deflected to Garland for a Republican.
The 0% Column: Five response types were deployed exclusively against Democrats: REFUSED, PERSONAL_ATTACK, DEFLECT_GARLAND, NO_RESPONSE, COUNTER_QUESTION. These represent 51% of all Democratic responses. Bondi has a separate toolkit she only unpacks for one party.
4. DEFLECTION TAXONOMY: WHERE QUESTIONS GO TO DIE
Deflection Types (n=39 exchanges with non-NONE deflection)
Deflection Signatures
DISTRICT_CRIME (10x) — The Primary Weapon
Pattern: Democratic member asks about Epstein/ICE killings → Bondi reads crime statistics from that member's home district
"He doesn't want to talk about what's happening in California... 4,561 criminal aliens released" (re: Lieu)
"She doesn't want to talk about crime in Atlanta... Christopher Welch crossed state lines to molest a 14-year-old" (re: McBath)
"He doesn't want to talk about his own district... Chase Mulligan preyed on girls in online chatrooms" (re: Raskin)
"Not even worth getting into the details" (re: Kamlager-Dove/Culver City)
Function: Converts oversight question into campaign attack ad. Forces member into defensive posture. Consumes their remaining time. Every single instance was pre-researched... Bondi had a dossier on each Democratic member's district.
GARLAND (7x) — Temporal Displacement
Pattern: Question about current DOJ conduct → "Did you ask Merrick Garland that?"
"I don't believe you asked Merrick Garland these questions when he was attorney general" (to Lieu re: Prince Andrew)
"None of them asked Garland one word about Epstein" (opening broadside)
Function: Procedural tu quoque. Shifts temporal frame from present (where Bondi has power and responsibility) to past (where she has neither). Structurally identical to "whataboutism" but dressed in institutional language.
TRUMP_PRAISE (5x) — Loyalty Display
Pattern: Any question → redirect to Trump accomplishments
"California is right up there with fraud. Donald Trump is coming to the rescue."
"He is the commander in chief... Trump is keeping this country safe, keeping this world safe."
Function: Signals alignment to audience. Converts testimony into endorsement. Most frequently deployed after crime stats questions from Republicans... a cooperative performance rather than a deflection.
THEATRICS (4x) + VICTIM_CLAIM (2x)
Pattern: When pressed → accuse questioner of bad faith or claim victimhood
"People who ask questions and don't want answers are doing theatrics"
"I'm not going to commit to anything to you because you won't let me answer"
"You don't get to reclaim your time because I don't answer to your satisfaction"
Function: DARVO (Deny, Attack, Reverse Victim and Offender). The witness reframes oversight as persecution, making the questioner the aggressor and herself the victim. Procedurally brilliant because it makes the act of questioning appear adversarial.
Deflection as Strategy, Not Accident: The deflection types are not random evasions. They form a coordinated tactical system: DISTRICT_CRIME punishes the questioner, GARLAND shifts blame backward, TRUMP_PRAISE rewards allies, THEATRICS delegitimizes the process, and VICTIM_CLAIM reverses the power dynamic. Each serves a distinct strategic function. This is preparation, not improvisation.
5. ATTACK MAP: PERSONAL ATTACKS AS WEAPON
All Personal Attacks by Bondi (n=26)
| Target | Party | Topic That Triggered It | Attack Content |
|---|
| Jayapal | D | EPSTEIN | "This is theatrics" |
| Nadler | D | EPSTEIN | "You were one of the leads on impeachment. Have you apologized?" |
| Lofgren | D | TRUMP_EPSTEIN | "How much money did she take from Reid Hoffman?" |
| Johnson | D | EPSTEIN | "People who ask questions and don't want answers" |
| Lieu | D | TRUMP_EPSTEIN | "Don't you ever accuse me of a crime. He has TDS." |
| Scanlon | D | DOM_TERRORISM | "You don't get anything regarding public safety. Nothing." |
| Scanlon | D | DOM_TERRORISM | "I'm not going to commit to anything to you" |
| Massie | R | EPSTEIN | "This guy has TDS. You're a failed politician." |
| Massie | R | EPSTEIN | "He voted against the deep fake AI porn ban. Hypocrite." |
| Raskin | D | EPSTEIN | "Do you even know who Chase Mulligan is? TDS." |
| Raskin | D | EPSTEIN | "Tiny little district... didn't even know who [predator] was" |
| Raskin | D | EPSTEIN | Attacked district crime instead of answering task force question |
| McBath | D | ICE_KILLINGS | Pivoted to Christopher Welch in Georgia |
| McClintock (time) | R | CRIME_STATS | Used his time to attack McBath's district |
| Balint | D | TRUMP_EPSTEIN | "Shame on you. This is pathetic." |
| Balint | D | OTHER | Anti-semitism accusation to Jewish member whose grandfather died in Holocaust |
| Ross | D | MAXWELL | "That's why you don't want to talk about Ireina Zerutka" |
| Goldman | D | EPSTEIN | "You're about as good a lawyer as when you tried to impeach Trump" |
| Goldman | D | EPSTEIN | "Don't yell at me" |
| Kamlager-Dove | D | DOM_TERRORISM | Attacked Culver City crime after time expired |
| Garcia/Quigley | D | WEAPONIZATION | Attacked chief of staff replacing him, Hakeem Jeffries Epstein money |
| Crockett | D | EPSTEIN | Attacked with TX illegal alien crime stats after time expired |
| Moscowitz | D | EPSTEIN | "Nothing is funny about mocking the Bible" |
| Neguse | D | CRYPTO | Attacked Boulder crime: "Muhammad Solomon... 82-year-old Holocaust survivor killed" |
| Bondi (general) | - | EPSTEIN | Called proceedings "theatrics" (opening + repeated) |
| Bondi (general) | - | CRIME_STATS | "Democrats don't want to talk about crime in their districts" |
Attack Trigger Analysis: Of 26 attacks, 16 (62%) were triggered by Epstein or Trump-Epstein questions. 4 (15%) by domestic terrorism questions. 3 (12%) by ICE killings. The remaining 3 by other topics. The attack function maps directly to threat level: the more dangerous the question to the administration, the more likely Bondi is to attack the questioner personally.
The "Burn Book" Pattern: Bondi came prepared with a dossier on each Democratic member... specific crime cases from their districts, past votes, campaign donations, personal history. This is not spontaneous anger. Multiple members referenced her "burn book." This represents hours of opposition research invested in deflecting questions before the hearing began.
Post-Time Attacks: In at least 8 instances, Bondi delivered personal attacks after a Democratic member's time had expired, using the next Republican's yielded time. This exploits a procedural gap: the attacked member cannot respond because their time is gone, but Bondi gets unlimited response time on borrowed Republican minutes. The attacked member's only recourse is a point of order (which Jordan consistently ruled against).
6. MEMBER SCORECARDS
Democrats — Effectiveness Ranking
| Member | St | Qs Asked | Responsive | Rate | Attacked? | Key Moment |
|---|
| Lieu | CA | 5 | 2 | 40% | 2x | Limo driver witness, "Don't accuse me of a crime" |
| Goldman | NY | 4 | 0 | 0% | 2x | Survivor hand-raise demonstration (most devastating moment) |
| Scanlon | PA | 4 | 0 | 0% | 2x | Domestic terror list... "secret list of Americans" |
| Raskin | MD | 3 | 0 | 0% | 3x | Emoluments, joint task force demand |
| Balint | VT | 5 | 0 | 0% | 2x | Lutnik/Failen/Fineberg Epstein ties |
| Nadler | NY | 3 | 0 | 0% | 1x | "How many co-conspirators indicted? Zero." |
| Jayapal | WA | 3 | 0 | 0% | 1x | Apology demand, time consumed by chair intervention |
| McBath | GA | 3 | 1 | 33% | 1x | Jordan Davis connection, ICE killings |
| Ross | NC | 4 | 2 | 50% | 1x | Maxwell prison privileges |
| Cohen | TN | 2 | 1 | 50% | No | Memphis task force, NIJ immigrant crime study |
| Neguse | CO | 3 | 1 | 33% | 1x | Jared Weise "kill him" video, PIN gutted to 2 people |
| Kamlager-Dove | CA | 3 | 1 | 33% | 1x | Hortman murders, DOJ scrubbed right-wing terror data |
| Lofgren | CA | 2 | 0 | 0% | 1x | Steve Tish emails, Trump name redaction |
| Crockett | TX | 0 | - | N/A | 1x | Refused to question Bondi, asked colleague instead |
| Garcia/Quigley | IL | 0 | - | N/A | 1x | "Worst AG in history" speech, impeachment demand |
Democratic Futility: The average Democratic response rate was 29%. But removing Lieu's anomalous death-threats exchange, it drops to ~22%. Multiple members (Raskin, Balint, Goldman, Scanlon, Nadler, Jayapal, Lofgren) achieved 0% responsive answers. The format structurally guarantees failure for adversarial questioners because: (a) 5 minutes is insufficient to force answers, (b) the chair controls procedural rulings, (c) "reclaim time" consumes the questioner's own clock.
Republicans — Effectiveness Ranking
| Member | St | Qs Asked | Responsive | Rate | Notable |
|---|
| Gill (TX) | TX | 17 | 17 | 100% | Rapid-fire Biden indictment. Most "productive" exchange. |
| Hageman | WY | 6 | 6 | 100% | Energy policy. All leading yes/no questions. |
| Fry (SC) | SC | 5 | 5 | 100% | Sanctuary cities. Rapid yes/no series. |
| Van Drew | NJ | 4 | 4 | 100% | Immigration. Cooperative Q&A. |
| Knott (NC) | NC | 8 | 8 | 100% | "True or false" series on crime stats. |
| Schmidt | KS | 4 | 4 | 100% | 287G, civil redress. Former AG. |
| McClintock | CA | 4 | 4 | 100% | Crime stats request. Cooperative. |
| Moscowitz | FL | 3 | 3 | 100% | Timeline facts. Then epic Epstein speech... attacked. |
| Massie | KY | 3 | 0 | 0% | ONLY Republican with 0%. Pressed on Wexner. Attacked 2x. |
| Jordan | OH | 5 | 4 | 80% | Chair. Controls flow. Brennan, Jack Smith, Lemon. |
The Massie Anomaly: Thomas Massie (R-KY) is the only Republican who received the "Democratic treatment"... 0% response rate, 2 personal attacks, including being called a "failed politician" and accused of "Trump derangement syndrome." His crime: pressing on Les Wexner's redaction as co-conspirator and demanding the prosecution/declination decisions. This proves the attack function is anti-accountability, not anti-Democrat. Any member who presses on Epstein co-conspirators triggers the defense mechanism regardless of party.
7. TEMPORAL ANALYSIS: BEHAVIOR OVER 4.5 HOURS
Hearing Phases (approximate)
| Phase | Time | Exchanges | Response Rate | Attacks | Character |
|---|
| Opening Statements | 0:00-0:30 | 3 | N/A | 0 | Scripted. Both sides set frames. |
| First Round (D/R) | 0:30-1:20 | 25 | ~55% | 4 | Bondi testing boundaries. First Garland deflections. |
| Second Round | 1:20-2:00 | 30 | ~50% | 6 | Escalation. Lieu limo driver, "Jackal and Hyde," survivor hand-raise #1. |
| Post-Recess | 2:00-2:35 | 25 | ~40% | 6 | Bondi more aggressive. Massie attacked. Raskin TDS x3. |
| Late Afternoon | 2:35-3:30 | 30 | ~35% | 6 | Fatigue + aggression spike. McBath, Ross, Balint targeted. Moscowitz epic timeline. |
| Final Hour | 3:30-4:31 | 39 | ~75% | 4 | Mostly Republican time. Rapid-fire Biden indictments. Goldman survivor demonstration. |
Fatigue Curve: Bondi's response rate decreased over the first 3.5 hours (from ~55% to ~35%) as she grew more aggressive and less responsive. But it spiked in the final hour because the last 45 minutes were dominated by cooperative Republican questioning (Gill rapid-fire, Knott true/false, Bonger). The format rewards the party that goes last.
Attack Escalation: Attacks were distributed roughly evenly across the middle sections (4-6 per phase), but intensified in severity over time. Early attacks were procedural ("theatrics"). By hour 3, Bondi was calling members "failed politicians," accusing them of "Trump derangement syndrome," making anti-semitism allegations to a Jewish member whose grandfather died in the Holocaust, and reading prepared district crime dossiers. The escalation pattern suggests deliberate calibration: start with deniable dismissiveness, escalate to personal destruction.
8. MATRIX: PARTY × RESPONSE TYPE
| DIRECT | FILIBUSTER | REFUSED | PARTIAL | PERSONAL_ATTACK | NO_RESPONSE | PENDING | DEFLECT_G | COUNTER_Q | TOTAL |
|---|
| DEM |
7 17% |
6 15% |
7 17% |
5 12% |
6 15% |
4 10% |
2 5% |
3 7% |
1 2% |
41 |
| GOP |
39 81% |
4 8% |
0 0% |
2 4% |
0 0% |
0 0% |
2 4% |
0 0% |
1 2% |
48 |
The Zero Row: Five cells in the GOP row are literally 0. Bondi never refused a Republican, never attacked one (as primary response), never deflected to Garland for one, never gave a Republican a non-response. The matrix is maximally asymmetric.
9. MATRIX: TOPIC × RESPONSIVE (Questions Only)
| Topic | Questions | Responsive | Non-Resp | Rate | Primary Asker Party |
|---|
| ENERGY | 6 | 6 | 0 | 100% | R |
| CHURCH | 4 | 4 | 0 | 100% | R |
| CRIME_STATS | 10 | 9 | 1 | 90% | R |
| SANCTUARY | 12 | 11 | 1 | 92% | R |
| IMMIGRATION | 10 | 9 | 1 | 90% | R mostly |
| FRAUD | 4 | 3 | 1 | 75% | R |
| WEAPONIZATION | 6 | 3 | 3 | 50% | Mixed |
| MAXWELL | 4 | 2 | 2 | 50% | D |
| DOJ_SURVEILLANCE | 3 | 2 | 1 | 67% | R |
| ICE_KILLINGS | 4 | 1 | 3 | 25% | D |
| EPSTEIN | 14 | 3 | 11 | 21% | D mostly |
| TRUMP_EPSTEIN | 5 | 0 | 5 | 0% | D |
| DOM_TERRORISM | 4 | 0 | 4 | 0% | D |
| CRYPTO | 2 | 0 | 2 | 0% | D |
| EMOLUMENTS | 1 | 0 | 1 | 0% | D |
Perfect Topic-Party Correlation: Every topic with 75%+ response rate is primarily asked by Republicans. Every topic with 25% or below is primarily asked by Democrats. The correlation coefficient between "% of questions from Republicans" and "response rate" is approximately r = 0.92. The topic itself may not matter... party of the questioner may be the sole determining variable.
The 0% Topics: Four topics received zero responsive answers: Trump-Epstein (5 questions), Domestic Terrorism (4 questions), Crypto (2 questions), and Emoluments (1 question). These represent the highest-threat topics to the administration. The complete refusal to engage is itself data... it marks the boundaries of what the executive branch considers unacceptable oversight.
10. MATRIX: TOPIC × DEFLECTION TYPE
| Topic | DISTRICT_CRIME | GARLAND | TRUMP_PRAISE | THEATRICS | BIDEN | VICTIM_CLAIM | OTHER |
|---|
| EPSTEIN | 4 | 3 | 1 | 2 | 1 | 2 | 2 |
| TRUMP_EPSTEIN | 2 | 0 | 2 | 0 | 0 | 0 | 1 |
| ICE_KILLINGS | 3 | 0 | 0 | 0 | 0 | 0 | 1 |
| DOM_TERRORISM | 1 | 0 | 0 | 0 | 0 | 0 | 2 |
| WEAPONIZATION | 0 | 0 | 0 | 1 | 2 | 1 | 1 |
| SANCTUARY | 0 | 0 | 1 | 0 | 0 | 0 | 1 |
| CRIME_STATS | 1 | 0 | 2 | 0 | 0 | 0 | 0 |
| IMMIGRATION | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Epstein is the Only Topic That Triggers ALL Deflection Types. It activated DISTRICT_CRIME, GARLAND, TRUMP_PRAISE, THEATRICS, BIDEN, VICTIM_CLAIM, and OTHER. No other topic triggered more than 3 types. Epstein is the nucleus of Bondi's defensive universe... every weapon in the arsenal gets deployed here.
DISTRICT_CRIME is the Dominant Counter-Move for Accountability Topics: It is the #1 deflection for Epstein (4x), ICE_KILLINGS (3x), and TRUMP_EPSTEIN (2x). When the question threatens the administration, the counter-move is always: attack the questioner's home constituency. This is the Resistance Ratchet in miniature... when M1 (answering) is blocked, M2 (attacking) activates, and the extraction of accountability-time is preserved.
11. MATRIX: UTTERANCE TYPE × RESPONSE TYPE
| Utterance | DIRECT | FILIBUSTER | REFUSED | PARTIAL | ATTACK | NO_RESP | PENDING | DEFLECT_G | COUNTER |
|---|
| YES_NO_Q | 30 | 5 | 4 | 3 | 2 | 0 | 2 | 1 | 0 |
| OPEN_Q | 13 | 3 | 3 | 3 | 3 | 1 | 2 | 2 | 1 |
| DEMAND | 2 | 1 | 1 | 2 | 2 | 4 | 0 | 0 | 0 |
| RHETORICAL | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
DEMANDS Are Least Effective: Of 12 demands, only 2 (17%) received direct answers. Demands triggered the most NO_RESPONSE (4) and PERSONAL_ATTACK (2). Demands signal maximum pressure but paradoxically produce minimum results because they give the witness justification to not answer ("I won't be ordered around").
YES_NO Questions Are Most Efficient: 30 of 47 (64%) received direct answers. But this is misleading... when filtered by party, GOP YES_NO questions had ~95% response rate while DEM YES_NO questions had ~30%. The question format only works when the witness wants to cooperate.
12. TRIPLE CROSS: PARTY × TOPIC × RESPONSIVE
Democrats Only — Response Rate by Topic
| Topic (Dem Questions) | Asked | Responsive | Rate |
|---|
| EPSTEIN | 10 | 1 | 10% |
| TRUMP_EPSTEIN | 5 | 0 | 0% |
| ICE_KILLINGS | 4 | 1 | 25% |
| DOM_TERRORISM | 4 | 0 | 0% |
| WEAPONIZATION | 4 | 1 | 25% |
| MAXWELL | 4 | 2 | 50% |
| CRYPTO | 2 | 0 | 0% |
| OTHER | 4 | 3 | 75% |
| IMMIGRATION | 2 | 2 | 100% |
Republicans Only — Response Rate by Topic
| Topic (GOP Questions) | Asked | Responsive | Rate |
|---|
| SANCTUARY | 12 | 11 | 92% |
| CRIME_STATS | 9 | 9 | 100% |
| IMMIGRATION | 8 | 7 | 88% |
| ENERGY | 6 | 6 | 100% |
| WEAPONIZATION | 4 | 4 | 100% |
| FRAUD | 4 | 3 | 75% |
| CHURCH | 4 | 4 | 100% |
| EPSTEIN | 4 | 2 | 50% |
| JACK_SMITH | 3 | 2 | 67% |
The Triple Cross Reveals the Mechanism: Even on the SAME topic, party determines outcome. On EPSTEIN, Democrats got 10% while Republicans got 50%. On WEAPONIZATION, Democrats got 25% while Republicans got 100%. On IMMIGRATION (small sample), Democrats actually got 100%... but that was Cohen's cooperative exchange. The topic is not the variable. The party is the variable. The topic just determines the magnitude of the gap.
13. STRUCTURAL DYSFUNCTION METRICS
Information Extraction Rate
37%
of all questions yielded NO useful information
33 of 89 questions/demands received non-responsive answers. If this were a deposition, opposing counsel would have sought court intervention after the third refusal.
Time Wasted on Deflection
~45 min
of 271 min total (opening + Q&A)
Estimated: 26 personal attacks (~30 sec each = 13 min), 10 filibusters (~1 min each = 10 min), 22 reclaim events (~30 sec each = 11 min), 7 Garland deflections (~1 min each = 7 min), misc (~4 min).
Binding Commitments Extracted
0
to Democratic members
Bondi made general promises ("we'll look at it," "we'll work with you") to Republicans but zero enforceable, specific commitments to any Democratic member. No documents promised, no deadlines set, no investigations committed to.
Contempt-Worthy Moments
8
exchanges where refusal arguably constituted contempt
Scanlon (domestic terror list), Goldman (prosecution memo), Balint (DOJ interviews of officials), Neguse (crypto team staffing), and 4 others where Bondi refused point-blank to provide factual information about DOJ operations.
Subpoena Threats Issued
0
by any member of either party
Despite 8+ contempt-worthy refusals, no member threatened subpoena, contempt proceedings, or any enforcement mechanism. The hearing had no teeth. Both sides knew it. Bondi knew it most of all.
New Information Produced
~3 items
previously unknown facts revealed
(1) Jared Weise confirmed employed at DOJ, (2) Goldman found 86-page prosecution memo still redacted, (3) Bondi confirmed "more indictments possible" on Epstein. Everything else was restated from prior public record.
Cost-Benefit Analysis: A 4.5-hour hearing with ~40 members, hundreds of staff hours, security, venue, and preparation... all to produce approximately 3 new facts, 0 binding commitments, 0 subpoenas, and ~45 minutes of campaign content for both sides. The information-per-hour yield approaches zero. This is the quantitative signature of accountability theater.
14. THE RESISTANCE RATCHET IN REAL TIME
EEDTM Application: The Resistance Ratchet predicts that when Mechanism 1 is blocked, elites shift to Mechanism 2 while preserving the extraction rate (Theta). Applied to congressional oversight:
M1: Direct Evasion (blocked by format)
The hearing format theoretically prevents simple refusal... members have 5 minutes, the witness is sworn in, cameras are rolling. But the 5-minute limit means M1 (silence) isn't necessary.
M2: Active Counter-Measures
When direct evasion might look bad, Bondi shifted to filibuster, personal attack, Garland deflection, district crime attack, etc. Each consumes the questioner's time while providing the appearance of engagement.
The Theta of Oversight: If we define "extracted accountability" as the ratio of binding commitments to questions asked, the Theta of this hearing is 0/89 = 0.00. If we use a looser definition (responsive answers / questions asked), Theta = 56/89 = 0.63. But 44 of those 56 responsive answers went to Republicans asking friendly questions. The accountability Theta for the minority party is 12/41 = 0.29... meaning the minority party "captured" only 29% of the answers it sought. The majority party captured 92%.
The Gamma of Oversight: Differential targeting ratio (Gamma) = GOP response rate / Dem response rate = 92/29 = 3.17x. For comparison, the subprime lending Gamma was 3.2x. Congressional oversight produces the same differential targeting ratio as predatory lending. The mechanism is different... instead of higher interest rates, it's lower information rates... but the mathematical signature is identical.
Resistance Ratchet Sequence Observed:
Step 1: Democrats ask Epstein questions (M1: direct accountability)
Step 2: Bondi deflects to Garland (shift to M2)
Step 3: Democrats reclaim time (M2 partially blocked)
Step 4: Bondi attacks member's district (shift to M3)
Step 5: Member's time expires (M3 succeeds)
Step 6: Next Republican yields time to Bondi (M4: post-time attack)
Result: Zero accountability extracted. Time consumed. Content produced for both bases. Theta preserved at 0.00.
15. THE BILATERAL DYSFUNCTION: SEVEN FRAMEWORKS FOR QUANTIFYING BOTH SIDES
Critical Concession: The preceding analysis focuses on Bondi's evasion. But intellectual honesty demands we acknowledge: the questioning was nearly as theatrical as the evasion. Democrats optimized for camera clips, not information extraction. Republicans optimized for cooperative performance, not oversight. Both sides practiced what Schopenhauer called eristic debate... competitive combat dressed as truth-seeking. Seven academic frameworks from rhetoric, linguistics, law, and political science allow us to quantify this bilateral dysfunction with formal definitions.
Framework 1: Aristotle's Three Genres of Rhetoric (c. 335 BCE)
Deliberative (symbouleutikon)
Audience: Decision-maker
Time: Future
Judgment: What is expedient?
Setting: Legislative assembly
"Should we reform this agency?"
Forensic (dikanikon)
Audience: Judge
Time: Past
Judgment: What is just?
Setting: Law court
"Did you obstruct this investigation?"
Epideictic (epideiktikon)
Audience: Spectator
Time: Present
Judgment: Who is virtuous?
Setting: Public ceremony
"The American people deserve to know..."
The Genre Collapse: The Bondi hearing was nominally forensic (investigating past actions) and occasionally deliberative (what should DOJ do next?). But it was functionally epideictic... both sides performing praise and blame for spectators who are not deciding anything. The audience that mattered was not in the room. When the audience shifts from decision-maker to spectator, rhetoric shifts from forensic/deliberative to epideictic regardless of the nominal setting. This is Aristotle's insight applied 2,300 years later.
Genre Classification of All 152 Exchanges
| Nominal Genre | Actual Genre | Count | % of Hearing | Who |
|---|
| Forensic | Forensic (genuine) | ~12 | 8% | Goldman docs, Massie redactions, Moscowitz timeline |
| Forensic | Epideictic (blame) | ~55 | 36% | Dem Epstein speeches, Crockett, Garcia, Quigley |
| Deliberative | Epideictic (praise) | ~48 | 32% | GOP crime stats, Biden indictment, Trump praise |
| Forensic | Eristic combat | ~26 | 17% | Personal attacks, burn book, TDS accusations |
| Deliberative | Deliberative (genuine) | ~11 | 7% | Hageman energy, Schmidt 287G, vaping |
Only 15% of the hearing was what it claimed to be. Genuine forensic exchanges (actually investigating facts) comprised ~8%. Genuine deliberative exchanges (actually deciding future action) comprised ~7%. The remaining 85% was epideictic performance or eristic combat. Both parties contributed roughly equally to the epideictic inflation.
Citation: Aristotle, Rhetoric, Book I, Chs. 3-8 (Kennedy trans., 2007, Oxford UP). See also Perelman & Olbrechts-Tyteca, The New Rhetoric (1969, Notre Dame UP) on "universal audience" vs. "particular audience."
Framework 2: Eristic vs. Dialectic (Schopenhauer, 1831)
Dialectic (dialektike)
Goal: Collaborative truth-seeking
Method: Both parties test propositions honestly
Outcome: Shared understanding
Origin: Plato's Socratic method
Eristic (eristike)
Goal: Victory regardless of truth
Method: Stratagems, tricks, emotional manipulation
Outcome: One party appears to win
Origin: Named for Eris, goddess of strife
Schopenhauer's 38 Stratagems Observed in Bondi Hearing
| # | Stratagem | Bondi Used | Questioners Used | Count |
|---|
| 1 | Extension (exaggerate opponent's claim, attack exaggeration) | "So you want NO oversight?" | "So you're covering up for ALL pedophiles?" | ~8 |
| 8 | Provocation (make opponent angry to cloud judgment) | Personal attacks, TDS accusations, burn book | Calling for resignation, "lied under oath" | ~14 |
| 18 | Interruption (break opponent's argument chain) | Filibuster through allotted time | Reclaim time before answer completes | ~22 |
| 26 | Retorsio (turn opponent's argument against them) | "Did YOU ask Garland that?" | "Your own office prosecuted Jordan Davis's killer" | ~9 |
| 28 | Ad Auditores (play to spectators, not opponent) | District crime stats for cameras | 4-minute speeches with 15-second questions | ~60+ |
| 36 | Bewilderment (overwhelm with meaningless verbosity) | Filibuster with tangent narratives | Compound multi-part questions | ~15 |
| 38 | Ad Personam (losing on merits → attack the person) | "Failed politician," "TDS," "hypocrite" | "Worst AG in history," "resign," "no decency" | ~26 |
Stratagem #28 (Ad Auditores) is the dominant move for BOTH sides. An estimated 60+ exchanges (40% of the hearing) were primarily addressed to spectators rather than to the nominal interlocutor. Democrats speechified for progressive media; Republicans performed for conservative media; Bondi performed for Trump. The hearing was a three-way eristic performance with no dialectical content. Schopenhauer's cynicism is vindicated: eristic is the natural mode, dialectic the rare exception.
Citation: Schopenhauer, A. (1831/1864). Die Kunst, Recht zu behalten (The Art of Being Right). See also Aristotle, Sophistical Refutations.
Framework 3: Irving Younger's 10 Commandments of Cross-Examination (1975)
The Courtroom Benchmark: Trial lawyers have formalized rules for extracting information from hostile witnesses. These rules work in depositions and courtrooms. Congressional questioners violate nearly all of them. This is not accidental... the format incentivizes violation. But the violation is quantifiable, and it measures the opportunity cost of performative questioning.
Younger's 10 Commandments Applied to the Bondi Hearing
| # | Commandment | Trial Standard | Hearing Reality | Violation Rate |
|---|
| 1 | Be brief | Short, focused questions | Avg Democratic "question" consumed 3-4 min of preamble before asking | ~85% |
| 2 | Plain words | Simple, clear language | Mostly compliant. Both sides used accessible language. | ~20% |
| 3 | Leading questions only | "You signed it, correct?" | Democrats used open-ended Qs ("Why did you...?") inviting filibuster | ~70% |
| 4 | Know the answer | Never ask blind | Goldman, Massie, Moscowitz knew answers. Others asked blind. | ~50% |
| 5 | Listen | Process answer before next Q | Reclaim time often cut off before processing. Some good listening (Lieu). | ~40% |
| 6 | Don't quarrel | Stay calm, let record speak | Outrage, arguing, "I believe you just lied under oath" | ~60% |
| 7 | No repetition | Once answered, move on | Same Epstein questions asked by 8+ different members | ~75% |
| 8 | Control explanations | Don't let witness speechify | Bondi filibustered through Democratic time repeatedly | ~80% |
| 9 | Limit questioning | Don't ask one too many | 5-minute format forces cramming. No strategic economy. | ~70% |
| 10 | Save for summation | Draw conclusions later | Every member tried to force conclusion in real time | ~90% |
Composite Cross-Examination Compliance Score: ~28/100 (weighted average of violation rates). A competent trial lawyer using Younger's rules and Pozner & Dodd's "Chapter Method" (one fact per question, leading questions, loops for contradictions) could have extracted an estimated 3-5x more information in the same time. The information extraction efficiency of congressional questioning is approximately 20-30% of courtroom cross-examination. The other 70-80% is consumed by speechifying, repetition, open-ended invitations to filibuster, and emotional quarreling that hands control to the witness.
The Structural Excuse: Congressional questioners lack tools trial lawyers have: (1) no judge to compel "yes or no" answers, (2) no contempt power with immediate teeth, (3) the 5-minute clock means the witness only needs to survive 300 seconds, (4) no sequential questioning (each member starts from zero). But the format itself was chosen by Congress. The 5-minute rule is self-imposed. The absence of enforcement mechanisms is a design choice. The format doesn't just fail to produce accountability... it is optimized to prevent it while maintaining the appearance of pursuing it.
Best Questioners by Cross-Examination Standards: Goldman (knew answers, had documents, used leading questions, controlled pace) and Massie (specific documents, one fact at a time, didn't speechify). Worst: Garcia/Quigley (pure speech, zero questions), Crockett (refused to question at all), early Jayapal (open-ended demand with no factual predicate). The correlation between cross-examination technique quality and information extraction is near-perfect.
Citations: Younger, I. (1976). "The Art of Cross-Examination," ABA Section of Litigation Monograph Series, No. 1. Pozner, L.S. & Dodd, R.J. (2004). Cross-Examination: Science and Techniques, 2nd ed., LexisNexis.
Framework 4: Grice's Cooperative Principle & Four Maxims (1975)
Quality
Don't say what you believe false. Don't say what you lack evidence for.
Quantity
Be as informative as required. Don't be more informative than required.
Manner
Avoid obscurity. Avoid ambiguity. Be brief. Be orderly.
Maxim Violations: Bondi vs. Questioners
| Maxim | Bondi Violations | Type | Questioner Violations | Type |
|---|
| Quality | "No evidence Trump committed a crime" (contested) | Violating (covert) | Presenting contested claims as established in preambles | Violating (covert) |
| Quantity (under) | "I'd have to check on that" / "Pending litigation" | Opting out | - | - |
| Quantity (over) | Filibuster: tangent narratives consuming clock | Violating | 4-minute speeches before 15-second questions | Violating |
| Relation | District crime stats when asked about Epstein | Flouting | Asking about topics for clip value, not case relevance | Flouting |
| Manner (obscurity) | Bureaucratic hedging: "to the extent that..." | Violating | Compound multi-part questions impossible to answer | Violating |
| Manner (prolixity) | Extended non-answer narratives | Violating | Verbose preambles consuming own time | Violating |
The Precondition Failure: Grice's framework presupposes shared communicative goals. The Cooperative Principle applies when both parties want the same thing from the conversation. In this hearing, goals are fundamentally misaligned: Bondi wants to survive without commitments; Democrats want viral clips of evasion; Republicans want viral clips of cooperation. If the "accepted purpose" is theater, then both sides are cooperating perfectly within that frame. The maxims aren't violated... they're operating under a different cooperative principle entirely: "Make your contribution such as is required for the production of compelling political spectacle."
Citation: Grice, H.P. (1975). "Logic and Conversation," in Cole & Morgan (Eds.), Syntax and Semantics, Vol. 3, pp. 41-58, Academic Press.
Framework 5: Speech Act Theory (Austin 1962 / Searle 1969)
The Illocutionary Gap: Surface Form vs. Actual Force
| Surface Grammar | Apparent Speech Act | Actual Illocutionary Force | Austin's Term | Examples |
|---|
| Question (interrogative) | Directive (seek info) | Assertive (accuse) | Pseudo-directive | "Isn't it true you covered up for pedophiles?" (not seeking info) |
| Question (interrogative) | Directive (seek info) | Expressive (outrage) | Pseudo-directive | "Don't the American people deserve better?" (rhetorical) |
| Answer (declarative) | Assertive (inform) | Null (evade) | Null-response commissive | "I appreciate the question" + filibuster (no info transferred) |
| Answer (declarative) | Assertive (inform) | Directive (counter-attack) | Inverted directive | "Did YOU ask Garland that?" (answer-as-question) |
| Promise (declarative) | Commissive (commit) | Null (empty) | Infelicitous commissive | "I'll look into it" (no mechanism, no deadline, no follow-up) |
| Statement (declarative) | Assertive (claim) | Performative (display) | Epideictic assertion | "I was a prosecutor for 18 years" (performing identity) |
The Illocutionary Gap Score: Of 89 questions/demands, an estimated ~52 (58%) were pseudo-directives... questions in grammatical form but assertions, expressions, or performances in illocutionary force. The questioner was not actually seeking information in most cases. Of Bondi's responses, an estimated ~28 (31%) were null-response commissives or inverted directives... answers in form but evasions, counter-attacks, or empty promises in force. The hearing's illocutionary gap (average mismatch between surface form and actual force) is approximately 45%. Nearly half of all utterances were not what they appeared to be.
Felicity Conditions (Austin): A question is "felicitous" (successful as a speech act) only if the speaker genuinely intends to receive and process new information. By this standard, fewer than 30 of 89 questions in the hearing were felicitous. The rest were infelicitous directives... they looked like questions, sounded like questions, but failed Austin's intentionality condition because the questioner wanted a clip, not an answer.
Citations: Austin, J.L. (1962). How to Do Things with Words, Oxford UP. Searle, J.R. (1969). Speech Acts, Cambridge UP. Searle (1976). "A Taxonomy of Illocutionary Acts," Minnesota Studies in Phil. of Science, 7, 344-369.
Framework 6: Pragma-Dialectics & the 10 Rules of Critical Discussion (van Eemeren, 1992)
Van Eemeren's 10 Rules — Violation Scorecard
| # | Rule | Bondi Violations | Questioner Violations | Severity |
|---|
| 1 | Freedom (don't prevent advancing/criticizing standpoints) | Refusing to engage topics; "I won't commit to anything to you" | Interrupting mid-answer; chair ruling against points of order | High |
| 2 | Burden of Proof (defend if asked) | "Pending litigation" on non-pending matters; blanket refusals | Presenting conclusions without evidentiary predicate | High |
| 3 | Standpoint (attack what was actually advanced) | Responding to questions not asked | Asking about topics the witness didn't raise | Moderate |
| 4 | Relevance (defend with relevant arguments) | District crime deflections; Garland deflections | Preambles unrelated to question | Critical |
| 5 | Unexpressed Premise (don't impute false premises) | "You don't care about victims" | "You're covering up for pedophiles" | Moderate |
| 6 | Starting Point (don't falsely claim shared premises) | "I think we can all agree Trump is keeping us safe" | "We all agree Epstein victims deserve justice" (while pursuing different goals) | Low |
| 7 | Argument Scheme (use appropriate reasoning) | False equivalence (Garland = Bondi) | Hasty generalization, slippery slopes | Moderate |
| 8 | Validity (logically valid reasoning) | Non sequiturs (Epstein Q → crime stats A) | Some non sequiturs in compound questions | Moderate |
| 9 | Closure (failed defense = retraction; successful = acceptance) | NEVER COMPLIED | NEVER COMPLIED | CRITICAL |
| 10 | Usage (clear, unambiguous formulations) | Bureaucratic hedging | Compound questions; rhetorical flourishes | Low |
Rule 9 is the Structural Kill Shot. Neither side ever conceded anything. No standpoint was retracted. No defense was accepted as successful. The concluding stage of critical discussion never arrives. This is not a failure... it is structurally impossible within the hearing format. A hearing with alternating 5-minute rounds, no shared burden of proof, no neutral arbiter, and no closure mechanism is, in van Eemeren's precise terminology, a pseudo-discussion: a rhetorical event wearing the formal clothing of a dialectical one.
Strategic Maneuvering (van Eemeren, 2010): Both sides simultaneously pursue dialectical reasonableness and rhetorical effectiveness. When rhetorical goals override dialectical obligations, the maneuvering "derails" and becomes fallacious. In this hearing, the derailment rate is approximately 85%... matching the Aristotelian epideictic classification above. The remaining 15% of genuine dialectical engagement occurred when both parties momentarily shared a goal (Lieu death threats, AI/cyber child exploitation, vaping).
Citations: van Eemeren & Grootendorst (1992). Argumentation, Communication, and Fallacies, Erlbaum. van Eemeren & Grootendorst (2004). A Systematic Theory of Argumentation, Cambridge UP. van Eemeren (2010). Strategic Maneuvering in Argumentative Discourse, Benjamins.
Framework 7: Stasis Theory (Hermagoras, c. 150 BCE / Cicero, 84 BCE)
1. Conjecture
"Did the act occur?"
Factual dispute
2. Definition
"What is it called?"
Classification dispute
3. Quality
"Was it justified?"
Justification dispute
4. Procedure
"Is this the right forum?"
Jurisdictional dispute
Stasis Mismatches in Key Exchanges
| Exchange | Questioner's Stasis | Bondi's Stasis | Match? | Productive? |
|---|
| Nadler: "How many co-conspirators indicted?" | Conjecture (fact) | Procedure (prior admin) | NO | NO |
| Lieu: "Were there underage girls at Trump-Epstein parties?" | Conjecture (fact) | Definition ("no evidence") | NO | NO |
| Scanlon: "Will you provide the domestic terror list?" | Conjecture (does list exist?) | Procedure ("won't commit") | NO | NO |
| Goldman: "Will you unredact the prosecution memo?" | Conjecture (will you?) | Procedure ("privileged") | NO | NO |
| Hageman: "Has DOJ filed lawsuits on energy EO?" | Conjecture (fact) | Conjecture (fact) | YES | YES |
| Van Drew: "Will removing criminal aliens save lives?" | Quality (justified?) | Quality (justified) | YES | YES |
| Gill: "Did Biden DOJ raid Trump's home?" | Conjecture (fact) | Conjecture (fact) | YES | YES |
| Balint: "Has DOJ asked Lutnik about Epstein ties?" | Conjecture (fact) | Procedure ("he addressed it himself") | NO | NO |
Stasis Alignment Score: Of the exchanges sampled, Democratic questions produced a stasis mismatch approximately 85% of the time. Republican questions produced a stasis match approximately 90% of the time. Bondi's primary defensive technique is stasis retreat: when asked a Conjecture question (did X happen?), she drops to Procedure (this isn't the right forum / prior administration) or Definition (I wouldn't characterize it that way). Stasis retreat is impossible when questioner and witness share an interest... which is why Republican questions produce matches. The mismatch is a feature of adversarial intent, not poor technique.
Citations: Cicero (c. 84 BCE). De Inventione (Hubbell trans., 1949, Loeb). Quintilian (c. 95 CE). Institutio Oratoria (Butler trans., 1920-22, Loeb). Fahnestock & Secor (1988). "The Stases in Scientific and Literary Argument," Written Communication, 5(4).
Framework Synthesis: The Bilateral Dysfunction Index (BDI)
Composite Bilateral Dysfunction Score
| Metric | Bondi Score | Questioner Score | Combined |
|---|
| Aristotelian Genre Fidelity (% forensic/deliberative) | 15% | 15% | 15% |
| Eristic vs. Dialectic (% dialectical) | ~10% | ~15% | ~12% |
| Cross-Exam Compliance (Younger score) | N/A (witness) | 28/100 | 28/100 |
| Gricean Cooperation (% maxim-compliant) | ~25% | ~30% | ~28% |
| Illocutionary Fidelity (form = force) | ~55% | ~42% | ~48% |
| Pragma-Dialectic Rule Compliance | ~15% | ~20% | ~18% |
| Stasis Alignment Rate | ~45% | ~50% | ~48% |
| BILATERAL DYSFUNCTION INDEX | 27/100 | 29/100 | BDI = 28/100 |
BDI = average of 7 framework compliance scores. 100 = perfect dialectical exchange. 0 = pure theater. The hearing scores 28/100... meaning 72% of the communicative activity was non-functional relative to its stated purpose. Critically, the scores are nearly symmetrical: Bondi's dysfunction score (27) and the questioners' dysfunction score (29) are within 2 points. The theater is bilateral.
The Central Finding: When seven independent academic frameworks from rhetoric (Aristotle), philosophy (Schopenhauer, Plato), linguistics (Austin, Searle, Grice), argumentation theory (van Eemeren), classical rhetoric (Hermagoras, Cicero), and legal practice (Younger, Pozner-Dodd) are applied to the same hearing, they converge on the same conclusion: the dysfunction is bilateral and nearly symmetrical. The witness evades at 73% dysfunction. The questioners perform at 71% dysfunction. The delta is 2 points. The hearing is not a case of one side failing... it is a structural equilibrium where both sides have optimized for theater over substance, and the format rewards that optimization.
Political Science Confirmation: This bilateral equilibrium is exactly what the institutional literature predicts. Frances Lee's Insecure Majorities (2016) documents how permanent competitive parity transforms Congress from governing institution to campaign platform. McCubbins & Schwartz (1984) showed "fire alarm" oversight selects for political salience over institutional importance. Mann & Ornstein's The Broken Branch (2006) and It's Even Worse Than It Looks (2012) chart the decline of oversight quality. Kriner & Schickler (2016) confirm investigations are driven by partisan incentives, not institutional logic. The Bondi hearing is not an outlier... it is the mathematical median of American oversight in 2026.
Additional citations: Lee, F.E. (2016). Insecure Majorities, U Chicago Press. McCubbins & Schwartz (1984). "Congressional Oversight Overlooked," AJPS 28(1). Mann & Ornstein (2006). The Broken Branch, Oxford UP. Mann & Ornstein (2012). It's Even Worse Than It Looks, Basic Books. Kriner & Schickler (2016). Investigating the President, Princeton UP. Aberbach (1990). Keeping a Watchful Eye, Brookings.
16. IMPLICATIONS FOR THE PAPER
"Accountability Theater: The Mathematical Structure of Congressional Dysfunction"
Thesis: Congressional oversight hearings are not dysfunctional by accident. They are structurally designed to produce the appearance of accountability while extracting near-zero actual accountability. The Bondi hearing demonstrates this with mathematical precision: a 64-point party differential in responsiveness, zero binding commitments, zero subpoenas, and an Accountability Theater Index of 78.4/100.
Key Data Points for the Paper:
- n = 152 coded exchanges across 4.5 hours
- 89 questions/demands, of which 56 (63%) received responsive answers
- 92% vs 29% party differential (p < 0.0001)
- Gamma = 3.17x (identical to subprime lending differential)
- 26 personal attacks, 92% against minority party
- 10 district-crime deflections with pre-researched dossiers
- 7 Garland deflections (temporal displacement)
- 0 binding commitments to minority party
- 0 subpoenas despite 8+ contempt-worthy refusals
- ~3 new facts produced in 271 minutes
- ATI = 78.4/100 (theater territory)
- The Massie Exception: Only R who pressed on Epstein → received D treatment (0% response, 2 attacks)
- Survivor demonstration: 0/all contacted by DOJ despite all reaching out
EEDTM Framework Connection: The hearing is a micro-demonstration of the same extraction mathematics documented across 20 BARSS cases. The extraction target is not money but accountability. The differential targeting (Gamma = 3.17x) maps to the same range as economic extraction. The Resistance Ratchet operates identically: when one evasion mechanism is blocked, the witness shifts to the next while preserving the extraction rate. Congressional oversight is not a check on power. It is a mechanism through which power performs its own legitimacy.
Comparative Frame: If Haiti's 1825 indemnity had a Theta of 0.86 (86% captured by elites), and this hearing had an accountability Theta of 0.00-0.29 depending on definition, then congressional oversight is actually less effective at extracting accountability from power than Haiti was at extracting wealth from the enslaved. The mechanism designed to constrain power produces less result than the mechanism designed to exploit the powerless.
BARSS Forensic Analysis | Rubik's Cube Cross-Analysis | Generated 2026-02-12
Source: Bondi House Judiciary Hearing Transcript, Feb 11, 2026 | 7,128 lines | 152 coded exchanges
Methodology: Manual forensic coding of every exchange with 14-variable schema
Wesley Bertil | BARSS Research Sciences
BONDI HEARING: MEMBER EFFECTIVENESS GRADES
Empirically Graded Using 6 Academic Frameworks | 152 Exchanges | 30+ Members Scored | Forensic, Not Emotional
THE RUBRIC: QUESTIONER EFFECTIVENESS RATING (QER)
Methodology: Each questioner is scored on 6 dimensions, each 1-10, for a maximum of 60. The dimensions are drawn from six independent academic traditions spanning 2,300 years of rhetorical, linguistic, and legal scholarship. Scores are converted to letter grades (A/B/C/D/F) and percentages. The same rubric is applied to both parties. This is not opinion... it is structural analysis using published criteria.
6 Dimensions of Questioner Effectiveness
1. Information Yield (IY)Shannon (1948)Did the questioning produce information not previously in the public record? 1 = zero new facts; 10 = multiple new facts extracted from a hostile witness.
2. Technique (TQ)Younger (1976) / Pozner-Dodd (2004)Compliance with cross-examination best practices: leading questions, one fact per question, brevity, control of explanations, no quarreling, no repetition. 1 = all commandments violated; 10 = textbook cross.
3. Illocutionary Fidelity (IF)Austin (1962) / Searle (1969)Were questions genuine directives seeking information, or pseudo-directives (speeches in question form)? 1 = pure speech, no genuine questions; 10 = every utterance genuinely sought new information.
4. Stasis Precision (SP)Hermagoras (c.150 BCE) / Cicero (84 BCE)Were questions asked at the correct level of dispute (conjecture/definition/quality/procedure)? 1 = total stasis mismatch; 10 = every question at the productive stasis level for its subject.
5. Argumentative Structure (AS)Toulmin (1958)Were claims evidence-backed with clear warrant? Complete Toulmin chain = Claim + Data + Warrant + Backing. 1 = bare assertions; 10 = every claim supported by documentary evidence with explicit reasoning.
6. Economy (EC)Grice (1975), Maxim of MannerTime efficiency. Ratio of question time to preamble time. Brevity. Avoidance of unnecessary prolixity. 1 = 95%+ preamble, 5% question; 10 = lean, direct, minimal setup.
Witness Accountability Rating (WAR) — Scored per interaction
1. Responsiveness (RS)Grice Maxim of RelationDid the answer address what was actually asked? 1 = completely non-responsive; 10 = directly answered every question.
2. Information Density (ID)Shannon (1948)New facts per response. 1 = zero information transferred; 10 = substantial new information provided.
3. Good Faith (GF)Van Eemeren Rules 1, 4 (1992)Absence of personal attacks, deflections, filibuster. 1 = pure bad faith; 10 = fully cooperative engagement.
4. Stasis Fidelity (SF)Hermagoras / CiceroAnswered at the same stasis level as the question. 1 = pure stasis retreat; 10 = engaged at questioner's stasis level.
5. Commitment Quality (CQ)Searle Commissives / Felicity ConditionsConcrete, enforceable promises with mechanisms. 1 = zero commitments or empty promises; 10 = specific, binding, time-bound commitments.
Grade Scale
| Score Range | Letter | Meaning |
|---|
| 83-100% | A | Effective cross-examination / genuine accountability |
| 67-82% | B | Above average... produced some results despite format constraints |
| 50-66% | C | Average... mixed technique, partial results |
| 33-49% | D | Below average... more performance than investigation |
| 0-32% | F | Pure theater... zero information yield, all epideictic |
FULL SCORECARD: ALL MEMBERS
Democrats — Ranked by QER
| Rank | Member | St | Qs |
IY
Shannon |
TQ
Younger |
IF
Austin |
SP
Stasis |
AS
Toulmin |
EC
Grice |
QER
/60 |
% |
Grade |
WAR
/50 |
WAR% |
| 1 | Goldman (NY) | NY | 4 | 8 | 8 | 8 | 7 | 9 | 7 | 47 | 78% | B+ | 5 | 10% |
| 2 | Massie (KY) R* | KY | 3 | 7 | 7 | 9 | 7 | 9 | 7 | 46 | 77% | B+ | 8 | 16% |
| 3 | Neguse (CO) | CO | 3 | 6 | 6 | 7 | 6 | 7 | 6 | 38 | 63% | C+ | 12 | 24% |
| 3 | Lieu (CA) | CA | 5 | 6 | 6 | 7 | 6 | 8 | 5 | 38 | 63% | C+ | 12 | 24% |
| 5 | Ross (NC) | NC | 4 | 5 | 6 | 7 | 7 | 6 | 6 | 37 | 62% | C | 19 | 38% |
| 5 | Balint (VT) | VT | 5 | 4 | 6 | 8 | 6 | 6 | 7 | 37 | 62% | C | 5 | 10% |
| 7 | Scanlon (PA) | PA | 4 | 4 | 5 | 7 | 6 | 6 | 5 | 33 | 55% | C | 5 | 10% |
| 7 | McBath (GA) | GA | 3 | 3 | 5 | 7 | 6 | 7 | 5 | 33 | 55% | C | 11 | 22% |
| 9 | Lofgren (CA) | CA | 2 | 4 | 4 | 6 | 5 | 7 | 4 | 30 | 50% | D+ | 7 | 14% |
| 9 | Cohen (TN) | TN | 2 | 3 | 5 | 6 | 5 | 6 | 5 | 30 | 50% | D+ | 19 | 38% |
| 9 | Raskin (MD) | MD | 3 | 2 | 4 | 6 | 7 | 7 | 4 | 30 | 50% | D+ | 5 | 10% |
| 9 | Kamlager-Dove (CA) | CA | 3 | 3 | 5 | 6 | 5 | 6 | 5 | 30 | 50% | D+ | 10 | 20% |
| 13 | Johnson (GA) | GA | 5 | 3 | 5 | 6 | 5 | 5 | 4 | 28 | 47% | D | 13 | 26% |
| 14 | Nadler (NY) | NY | 3 | 3 | 4 | 5 | 6 | 6 | 3 | 27 | 45% | D | 6 | 12% |
| 15 | Jayapal (WA) | WA | 3 | 2 | 3 | 4 | 5 | 5 | 3 | 22 | 37% | D- | 8 | 16% |
| 16 | Crockett (TX) | TX | 0 | 1 | 1 | 2 | 3 | 5 | 3 | 15 | 25% | F | 3 | 6% |
| 17 | Garcia/Quigley (IL) | IL | 0 | 1 | 1 | 1 | 2 | 4 | 2 | 11 | 18% | F | 2 | 4% |
*Massie Exception: Thomas Massie (R-KY) is ranked with Democrats because he received the "Democratic treatment" from Bondi (0% responsive, 2 personal attacks). His QER of 77% makes him the #2 most effective questioner of the hearing regardless of party. He is the only member on either side who combined document-based evidence (Toulmin: 9) with genuine information-seeking questions (Illocutionary Fidelity: 9). His "reward" was being called a failed politician.
Republicans — Ranked by QER
| Rank | Member | St | Qs |
IY
Shannon |
TQ
Younger |
IF
Austin |
SP
Stasis |
AS
Toulmin |
EC
Grice |
QER
/60 |
% |
Grade |
WAR
/50 |
WAR% |
| 1 | Jordan (OH) | OH | 5 | 5 | 7 | 7 | 7 | 7 | 7 | 40 | 67% | B- | 34 | 68% |
| 2 | Schmidt (KS) | KS | 4 | 4 | 7 | 7 | 7 | 7 | 7 | 39 | 65% | C+ | 35 | 70% |
| 3 | Harris (NC) | NC | 3 | 4 | 6 | 7 | 7 | 7 | 6 | 37 | 62% | C | 36 | 72% |
| 4 | Moscowitz (FL) | FL | 3 | 5 | 6 | 6 | 6 | 8 | 4 | 35 | 58% | C | 26 | 52% |
| 4 | Cline (VA) | VA | 3 | 4 | 6 | 7 | 6 | 6 | 6 | 35 | 58% | C | 31 | 62% |
| 4 | Palmer (AL) | AL | 2 | 4 | 6 | 7 | 6 | 6 | 6 | 35 | 58% | C | 32 | 64% |
| 7 | Biggs (AZ) | AZ | 4 | 3 | 6 | 7 | 6 | 6 | 6 | 34 | 57% | C | 24 | 48% |
| 7 | Hageman (WY) | WY | 6 | 3 | 7 | 5 | 7 | 5 | 7 | 34 | 57% | C | 38 | 76% |
| 9 | Issa (CA) | CA | 1 | 3 | 6 | 6 | 7 | 5 | 5 | 32 | 53% | C | 34 | 68% |
| 9 | Tiffany (WI) | WI | 3 | 4 | 6 | 6 | 6 | 5 | 5 | 32 | 53% | C | 33 | 66% |
| 9 | Kiley (CA) | CA | 3 | 3 | 6 | 6 | 6 | 5 | 6 | 32 | 53% | C | 33 | 66% |
| 9 | Fry (SC) | SC | 5 | 2 | 7 | 4 | 7 | 5 | 7 | 32 | 53% | C | 35 | 70% |
| 13 | Knott (NC) | NC | 8 | 2 | 8 | 3 | 7 | 3 | 8 | 31 | 52% | C | 33 | 66% |
| 14 | Gill (TX) | TX | 17 | 1 | 8 | 2 | 7 | 3 | 9 | 30 | 50% | D+ | 31 | 62% |
| 14 | Salazar (FL) | FL | 2 | 3 | 5 | 6 | 6 | 5 | 5 | 30 | 50% | D+ | 32 | 64% |
| 16 | Van Drew (NJ) | NJ | 4 | 2 | 6 | 4 | 7 | 5 | 5 | 29 | 48% | D | 35 | 70% |
| 17 | Andre (MO) | MO | 1 | 2 | 5 | 5 | 6 | 5 | 5 | 28 | 47% | D | 30 | 60% |
| 17 | McClintock (CA) | CA | 4 | 3 | 5 | 4 | 6 | 4 | 5 | 27 | 45% | D | 32 | 64% |
| 17 | Bonger (FL) | FL | 2 | 2 | 5 | 5 | 6 | 4 | 5 | 27 | 45% | D | 33 | 66% |
AGGREGATE SCORES: THE BIG PICTURE
Democratic QER Average
51%
Grade: C (30.7/60)
Range: 18% (Garcia) to 78% (Goldman). High variance. Best were genuine investigators. Worst were pure speechmakers.
Republican QER Average
54%
Grade: C (32.1/60)
Range: 45% (Bonger) to 67% (Jordan). Low variance. Uniformly cooperative, uniformly confirmatory.
Combined QER Average
52%
Grade: C (31.5/60)
The entire committee averaged a C. Both parties failed to produce effective oversight, through different mechanisms.
Bondi WAR vs Democrats
17%
Grade: F (8.5/50)
Average across all Dem interactions. Near-total non-cooperation. Personal attacks, stasis retreat, zero commitments.
Bondi WAR vs Republicans
66%
Grade: C+ (33.0/50)
Average across all GOP interactions. High responsiveness but low information density... she answered but said nothing new.
WAR Differential (Gamma)
3.9x
GOP gets 3.9x the witness cooperation
On the WAR scale (which includes good faith + commitments), the differential is even larger than the raw responsive rate (3.17x).
THE DIMENSION ANALYSIS: WHERE EACH SIDE FAILS
Average Score per Dimension — Democrats vs Republicans
| Dimension | Source | Dem Avg | GOP Avg | Delta | Interpretation |
|---|
| Information Yield | Shannon | 3.5 | 3.0 | +0.5 D | Neither party extracted much. Dems slightly higher because hostile questioning occasionally forces revelations through refusal patterns. |
| Technique | Younger | 4.8 | 6.2 | +1.4 R | Republicans used better technique (leading Qs, brevity) because cooperative witnesses reward technique. Hostile witnesses punish it. |
| Illocutionary Fidelity | Austin/Searle | 5.8 | 5.1 | +0.7 D | Democrats asked more genuine questions. Republicans asked more confirmatory pseudo-directives. |
| Stasis Precision | Hermagoras | 5.5 | 6.5 | +1.0 R | GOP questions matched Bondi's stasis level because they shared interests. Dem questions hit stasis walls. |
| Argumentative Structure | Toulmin | 5.9 | 5.2 | +0.7 D | Democrats brought more documentary evidence (Toulmin "data"). Republicans relied on shared premises. |
| Economy | Grice Manner | 4.3 | 6.0 | +1.7 R | BIGGEST GAP. Democrats spent ~70% of time on preambles. Republicans were leaner. The speech habit is the single biggest Democratic weakness. |
The Democratic Paradox: Democrats scored higher on Illocutionary Fidelity (+0.7) and Argumentative Structure (+0.7) and Information Yield (+0.5)... meaning they asked more genuine questions, with better evidence, and extracted slightly more new information. But they scored lower on Technique (-1.4), Stasis Precision (-1.0), and especially Economy (-1.7). They asked better questions badly. The speech habit consumed their own time, leaving Bondi less to fill with filibuster... but also leaving themselves less time for follow-up.
The Republican Paradox: Republicans scored higher on Technique (+1.4), Stasis (+1.0), and Economy (+1.7)... meaning they were technically better questioners. But they scored lower on Illocutionary Fidelity (-0.7) and Argumentative Structure (-0.7)... meaning their questions were less genuine and less evidence-backed. They asked worse questions well. Perfect technique applied to confirmatory pseudo-directives produces procedurally flawless theater.
The Synthesis: Both parties average a C (52%). Democrats fail on execution. Republicans fail on substance. The result is identical: a C-grade hearing that produces no accountability. This is the bilateral equilibrium. The format rewards Democratic speechmaking (clips) and Republican confirmation (talking points) equally, while punishing the only behavior that would produce accountability: Goldman/Massie-style document-based, technique-compliant, genuine interrogation.
TOP 5 AND BOTTOM 5: ALL MEMBERS, BOTH PARTIES
Top 5 Most Effective Questioners
| # | Member | Party | QER% | Grade | Why |
|---|
| 1 | Goldman | D | 78% | B+ | Documents. Leading Qs. Survivor demonstration. Found 86-pg prosecution memo. |
| 2 | Massie | R | 77% | B+ | Co-conspirator document. One fact at a time. Wexner redaction exposed. |
| 3 | Jordan | R | 67% | B- | Chair advantage. Leading Qs. Brennan referral confirmed. Surveillance policy. |
| 4 | Schmidt | R | 65% | C+ | Former AG. Precise legal questions. 287G, Rule 65 bond requirement. |
| 5 | Neguse | D | 63% | C+ | Weise video. PIN staffing exposed (35→2). Crypto team elimination. |
Bottom 5 Least Effective Questioners
| # | Member | Party | QER% | Grade | Why |
|---|
| 31 | Garcia/Quigley | D | 18% | F | Zero questions. Pure speech. Called for resignation. Zero information. |
| 30 | Crockett | D | 25% | F | Refused to question witness. Asked colleague instead. Performance piece. |
| 29 | Jayapal | D | 37% | D- | Open-ended demand. Lost control to chair. Time consumed by procedure. |
| 28 | Bonger | R | 45% | D | Open-ended. "Brag about ATF." Softball. Zero new information. |
| 27 | McClintock | R | 45% | D | "Repeat the crime stats." Purely confirmatory. Yielded time for attack on McBath. |
The Goldman-Massie Standard: Only 2 of 30+ members scored above 70%. Both brought documents. Both used leading questions. Both asked genuine questions (IF: 8-9). Both were attacked for it. Goldman and Massie represent what congressional oversight could be if members prioritized extraction over performance. Their technique compliance scores (8 and 7) and Toulmin scores (9 and 9) show that effective questioning is possible within the 5-minute format... it's just almost never attempted.
THE GILL PARADOX: PERFECT TECHNIQUE, ZERO INFORMATION
Rep. Gill (R-TX): 17 Questions in ~3 Minutes
| Dimension | Score | Explanation |
|---|
| Information Yield | 1 | Every answer was already public knowledge. Zero new facts. |
| Technique | 8 | Perfect yes/no format. Rapid. Controlled. Textbook Younger. |
| Illocutionary Fidelity | 2 | Not seeking information. Building a prosecutorial narrative re: Biden DOJ. |
| Stasis Precision | 7 | Correct stasis (conjecture) but trivially so... facts already established. |
| Argumentative Structure | 3 | No evidence presented. Relied on Bondi's confirmation as sole data. |
| Economy | 9 | Most time-efficient member of the hearing. Pure signal. |
QER: 30/60 = 50% (D+). Gill demonstrates that technique without substance is as theatrical as substance without technique. His exchange was the most "productive" by raw output (17 answers) and the least informative (0 new facts). This is the Potemkin village of oversight: structurally perfect, informationally empty.
Rep. Goldman (D-NY): 4 Questions, Most Devastating
| Dimension | Score | Explanation |
|---|
| Information Yield | 8 | Found 86-pg prosecution memo. Draft indictment. Survivor demo. |
| Technique | 8 | Specific documents. Leading questions. Controlled pace. |
| Illocutionary Fidelity | 8 | Genuine directives with documentary support. |
| Stasis Precision | 7 | Correct stasis. Forced engagement before Bondi could retreat. |
| Argumentative Structure | 9 | Documents as data. FVTPA as warrant. Complete Toulmin chains. |
| Economy | 7 | Efficient. Used time for evidence, not speeches. |
QER: 47/60 = 78% (B+). Goldman proves what's possible. 4 questions produced more new information than every other Democrat combined. The survivor hand-raise demonstration (0 contacted by DOJ / all reached out / all ignored / all still willing) was the hearing's single most devastating moment... and it required no cooperation from Bondi at all.
BONDI'S WITNESS ACCOUNTABILITY RATING: PER-MEMBER BREAKDOWN
WAR Score per Democratic Member (max 50)
| Member | RS
Grice | ID
Shannon | GF
Eemeren | SF
Stasis | CQ
Searle | WAR | % | Attacked? |
|---|
| Raskin | 1 | 1 | 1 | 1 | 1 | 5 | 10% | 3x |
| Scanlon | 1 | 1 | 1 | 1 | 1 | 5 | 10% | 2x |
| Balint | 1 | 1 | 1 | 1 | 1 | 5 | 10% | 2x |
| Goldman | 1 | 1 | 1 | 1 | 1 | 5 | 10% | 2x |
| Nadler | 1 | 1 | 2 | 1 | 1 | 6 | 12% | 1x |
| Lofgren | 2 | 1 | 1 | 2 | 1 | 7 | 14% | 1x |
| Jayapal | 2 | 1 | 2 | 2 | 1 | 8 | 16% | 1x |
| Kamlager-Dove | 3 | 2 | 2 | 2 | 1 | 10 | 20% | 1x |
| McBath | 3 | 2 | 2 | 3 | 1 | 11 | 22% | 1x |
| Lieu | 3 | 2 | 2 | 2 | 3 | 12 | 24% | 2x |
| Neguse | 3 | 3 | 3 | 2 | 1 | 12 | 24% | 1x |
| Johnson | 4 | 2 | 3 | 3 | 1 | 13 | 26% | 1x |
| Cohen | 5 | 3 | 5 | 4 | 2 | 19 | 38% | No |
| Ross | 5 | 4 | 3 | 5 | 2 | 19 | 38% | 1x |
The WAR Floor: Four Democrats (Raskin, Scanlon, Balint, Goldman) hit the absolute floor of 5/50 (10%)... meaning Bondi gave them effectively zero cooperation on every dimension. These are the members who pressed hardest on Epstein, domestic terrorism, and Trump-Epstein connections. The correlation between WAR score and topic threat level is nearly perfect. Bondi's cooperation is not random... it is a precise function of how threatening the question is to the administration.
The Cohen-Ross Anomaly: Cohen (38%) and Ross (38%) received notably higher WAR scores than other Democrats. Cohen asked about Memphis (a cooperative topic) and cited the NIJ study (non-threatening). Ross asked about Maxwell (where Bondi could say "she should die in prison" without cost). When Democrats ask questions that Bondi can answer without political risk, she cooperates. The WAR score measures Bondi's cost-benefit calculation, not the member's technique.
COMBINED EFFECTIVENESS: QER × WAR PRODUCT
The Product Score: Multiply QER% × WAR% to get the Combined Effectiveness (CE)... the probability that a given member's questioning produces both good questions AND substantive answers. A CE above 25% means at least some genuine accountability is occurring. Below 10% means the exchange is pure theater.
| Member | Party | QER% | WAR% | CE | Interpretation |
|---|
| Schmidt (KS) | R | 65% | 70% | 45.5% | Highest CE. Former AG asking precise legal Qs to cooperative witness. |
| Jordan (OH) | R | 67% | 68% | 45.6% | Chair + technique + cooperation. Highest combined efficiency. |
| Harris (NC) | R | 62% | 72% | 44.6% | Church questions. Genuine legal inquiry with cooperative witness. |
| Hageman (WY) | R | 57% | 76% | 43.3% | Energy. Highest WAR. Most cooperative interaction of the hearing. |
| Fry (SC) | R | 53% | 70% | 37.1% | Sanctuary yes/no series. Efficient if confirmatory. |
| Moscowitz (FL) | R | 58% | 52% | 30.2% | Mixed. Good factual Qs early, then epic speech tanked economy. |
| Ross (NC) | D | 62% | 38% | 23.6% | Best Dem CE. Maxwell topic allowed some cooperation. |
| Cohen (TN) | D | 50% | 38% | 19.0% | Memphis. Non-threatening topic. Some genuine exchange. |
| Lieu (CA) | D | 63% | 24% | 15.1% | Good technique destroyed by Bondi's refusal to engage. |
| Neguse (CO) | D | 63% | 24% | 15.1% | Same pattern. Good Qs, stonewalled. |
| Goldman (NY) | D | 78% | 10% | 7.8% | BEST questioner × WORST witness cooperation = near-zero. |
| Massie (KY) | R | 77% | 16% | 12.3% | Same. Documents + genuine Qs met with attacks. |
| Raskin (MD) | D | 50% | 10% | 5.0% | Ranking member vs. AG = pure theater. 3 attacks, 0 answers. |
| Balint (VT) | D | 62% | 10% | 6.2% | Good questions. Total stonewalling. Anti-semitism accusation. |
| Garcia/Quigley | D | 18% | 4% | 0.7% | The floor. Pure speech meets pure hostility. 0.7% combined. |
The Goldman Tragedy: The hearing's best questioner (QER: 78%) paired with the hearing's worst witness cooperation (WAR: 10%) produces a CE of 7.8%. Goldman did everything right by cross-examination standards and still extracted almost nothing because the format provides no enforcement mechanism. This is the structural proof that the format itself is the problem. No amount of questioner skill can overcome a witness who has decided not to cooperate and faces no consequences for refusal.
The Inverse Correlation: Plot QER against WAR for Democrats and you get a negative correlation. The better the questioning technique, the lower the witness cooperation. Goldman (78% QER → 10% WAR). Balint (62% QER → 10% WAR). Cohen (50% QER → 38% WAR). Garcia (18% QER → 4% WAR is an outlier). The relationship is: better questions = more threatening = less cooperation. The format punishes exactly the behavior it should reward.
IMPLICATIONS FOR THE PAPER
What the Grading System Proves:
- Both parties average a C (52%/54%). The dysfunction is bilateral, not one-sided.
- Democrats fail on Economy and Technique. They speechify instead of interrogating. The speech habit is their single biggest structural weakness (-1.7 vs GOP).
- Republicans fail on Illocutionary Fidelity and Information Yield. They ask fake questions to a cooperative witness. Perfect technique applied to pseudo-directives.
- Only 2 of 30+ members scored B+ or above. Both brought documents. Both used Younger-compliant technique. Both were attacked for it.
- The Combined Effectiveness (QER × WAR) reveals the structural trap: the best Democratic questioners produce lower CE scores than mediocre Republican ones because witness cooperation is a function of political alignment, not question quality.
- The format rewards the wrong behaviors: speechmaking (Dem), confirmation (GOP), evasion (witness). It punishes the only right behavior: evidence-based adversarial interrogation.
- Gamma on the WAR scale = 3.9x... even larger than the raw responsive rate gap (3.17x), because the WAR includes good faith and commitments.
Recommended Structural Reforms (from the data):
- Staff-led questioning: 30-minute blocks led by trained attorneys using Pozner-Dodd methodology. The Jan 6 Committee did this. It worked.
- Enforcement mechanism: Real-time contempt referrals for non-responsive answers, like a court's power to instruct "answer the question."
- Sequential examination: One examiner builds a complete Toulmin chain instead of 30 members each starting from zero.
- Document-first protocol: All documentary evidence entered before oral questioning begins, preventing the witness from claiming unfamiliarity.
- No yielded time: Eliminate the practice of yielding time to the witness for counter-attacks on prior questioners.
BARSS Forensic Analysis | Member Effectiveness Grades | Generated 2026-02-12
Rubric: 6-dimension QER (Shannon, Younger, Austin/Searle, Hermagoras/Cicero, Toulmin, Grice) + 5-dimension WAR (Grice, Shannon, van Eemeren, Hermagoras, Searle)
Source: Bondi House Judiciary Hearing Transcript, Feb 11, 2026 | 152 coded exchanges
Wesley Bertil | BARSS Research Sciences