Subscribe now

I checked my own numbers. Four of them were wrong.

Monday, I published The Silence. It reported that 167 people commented on two Facebook posts, and that not one of them argued the oversight architecture the essay was about.

Tuesday I sat down to review the workbook that backs it. I opened the file to trace every count in the essay to the row it came from.

Four counts had no row.

The essay said 33 people argued about the Secretary of Defense. Twenty-four about DEI. Fourteen about the video the six members of Congress made. Ten about the Commander in Chief’s authority. I could not find where any of those came from. I made a mistake. Two turned out to be code counts wearing a subject label. The rest, well.

So I read all 168 comments again, coded each one for subject, wrote down the boundary rules I was using, and counted.

The Secretary of Defense: 28. DEI: 16. The Commander in Chief’s authority: 8. Officer promotion mechanics: 20, which the essay never reported at all.

And the video: 30.

That last number is why I am telling you this instead of quietly fixing it. The count I got most wrong runs against me. More than twice as many readers argued the Congressional video as I reported. Part Two drew 40 comments, and 30 of them argued the video, the sedition charge, or whether an illegal order had been given.

Thirty people arguing hard about the military. Still, none of them arguing the four channels.

The finding got stronger when I checked it.

I corrected the essay Tuesday with a dated note. Every count in it now traces to a row you can open. This installment is that row, and the other 167.

Share

Why you are getting this free

Coded Threads is meant to be a monthly paid series. The paid tier at What Time Binds is not open yet. It opens when enough folks support my work.

Publishing installment №01 behind a wall that does not exist yet would hide the only thing that argues for the series. So take the whole thing. Memo, workbook, codebook, a blank template wired to the same scheme, and the raw threads I coded.

If you want this monthly, pledge. Pledges are what tell me it is time to turn the tier on.

Downloads

Six files. Take all of them.

1 · The memo. Everything below, as a PDF. Method, findings, what surprised me, what I would do differently, and the corrections.

Coded Threads 01 Memo

82.7KB ∙ PDF file

Download

Download

2 · The workbook. 168 anonymized turns across five tabs. Every count is a live formula, so recoding a row moves the totals. This is the file that found the four bad numbers.

Coded Threads 01 Workbook

51KB ∙ XLSX file

Download

Download

3 · The codebook. 42 codes across 8 families, plus 6 subject tags. Each code carries a definition, a rule for applying it, a rule for not applying it, an example, and its count.

Coded Threads 01 Codebook V1

97.1KB ∙ PDF file

Download

Download

4 · The blank template. The same scheme, empty. Dropdowns loaded, frequency tabs wired. Paste your own thread in, and it counts itself.

Coded Threads Blank Template

39.4KB ∙ XLSX file

Download

Download

5 · The quickstart. Five steps on one page. Capture, load, first pass, second pass, reconcile.

Coded Threads 01 Quickstart

39.7KB ∙ PDF file

Download

Download

6 · The raw threads. Both Facebook captures in one workbook, a tab per thread, 194 turns. Commenter names are replaced with the same P-01 through P-92 pseudonyms the coded workbook uses, so you can re-code the material without anyone being named. Paste the verbatim column into the blank template and tell me where we diverge.

Personal use, attribution required, no republishing. Method questions, corrections, divergent codings, and takedown requests: info@what-time-binds.com

Metadata

1. What was coded

Two Facebook posts from What Time Binds published four days apart on my personal page.

The first, Called It, went up Sunday morning, August 31, 2026. It named the five Secretaries of Defense who signed the February 27, 2025 letter warning Congress about the reckless firings of general officers, and it named what happened in the eighteen months since. The post drew 128 reader comments across four days.

The second, When Order Becomes Oracle, Part Two, went up Wednesday, September 3, 2026. It named the six members of Congress (all veterans or intelligence officers) who released a ninety-second video restating what the constitutional meaning of “order” actually is. The post drew 40 comments across three days.

I captured both threads on Saturday, September 6, by screenshotting the full comment section on each and transcribing every reader comment to a text file. What I coded was the reader turns only — my own comments as original poster stayed out of the coding, though they appear in the raw thread capture for reference.

The coding pass ran across the weekend. Two passes, one open and one focused, using a codebook I built for the first pass and refined for the second. Every count in the Monday essay traces to a row you can inspect in the workbook that ships with this installment.

2. The codebook, briefly

The framework is hybrid. It uses grounded-theory-style open coding for the first pass and a deductive-inductive scheme for the second, the same practice I used in my USC Rossier dissertation on media literacy during the COVID misinformation years.

Codes sort into eight families:

• A · Service-Anchor Moves — claims anchored to a service record, a family member’s service, a credential, or direct experience of how the institution works

• B · Precedent and History — arguments that place the claim in history, deployed as reassurance or as warning

• C · Frame Maneuvers — moves that shift the frame rather than answer inside it

• D · Loyalty and Civilian Control — the elected branch and the professional military, and the comply-or-resign binary

• E · Sedition and Specificity — what counts as sedition, what counts as an illegal order, and how specific a charge must be

• F · Fact and Source — sourcing, primary documents, correction, and what happens to a reader who gets corrected

• G · Slogan and One-Liner — positions delivered with no argument attached, and the replies they draw

• H · Moderation and Delivery — how a turn is delivered, insult wrapped around a real point above all

Forty-two codes sit inside those eight families. Forty-one were applied to this dataset; C4 was defined and never used. The full codebook ships as a separate PDF.

Subject is a second dimension, counted separately. A code names the move a turn makes. A subject names what the turn is about. Six subject tags were derived on September 9, 2026 by reading every respondent turn, with the boundary rules published in the codebook so the counts can be disputed.

Forty-one specific codes sit inside those eight families. The full codebook, with definitions, decision rules, and one example per code, ships as a separate PDF in this installment.

3. Findings

The workbook produced one finding above the others, and the Monday essay is built around it.

The essay was about oversight architecture. Nobody in the thread argued the architecture.

The essay named four channels that stop an unlawful order inside the U.S. military: Judge Advocate General review, Inspector General oversight, chaplaincy, and 10 U.S.C. § 1034, the Military Whistleblower Protection Act. It named what happened to each between January 2025 and September 2026. It named Adm. Alvin Holsey, pushed out of Southern Command two years early after raising legal concerns about the Caribbean boat strikes.

Zero comments across the 168 argued about the JAG-office removals. Zero argued about the seventeen Inspectors General dismissed on the night of Friday, January 24, 2025. Zero argued about chaplaincy replacements or Doug Wilson’s Pentagon invitation. Zero argued about § 1034. Zero named Holsey by name.

What readers argued about instead:

• The Congressional video from Part Two (30 comments)

• The Secretary of Defense as a person (28 comments)

• Officer promotion and selection board mechanics (20 comments)

• DEI and affirmative action (16 comments)

• The Commander in Chief’s constitutional authority (8 comments)

• Obama firing officers a decade ago (7 comments)

One hundred of the 168 turns carry at least one of those six subjects. Sixty-eight carry none: insult with no claim attached, bare agreement, thread housekeeping.

Every one of those is a real subject. None of them is the subject the essay named.

The second finding, which is what most surprised me in the coding, is about delivery.

Thirty of the 168 comments — 18 percent — carried a substantive counter argument delivered inside a personal attack. Not the top argument. Not the top valence. The top delivery mode. Nazi-POS. Hitler-dildo. Snuff-porn. Comments of that shape came from both sides.

More came from the ally side than the opposition side. That is a coalition problem worth naming, and the Monday essay names it. Every time an ally delivers a correct substantive counter with slur and class contempt, the opposition side gets to say “these people are unserious” with evidence to point at.

The third finding is that the two threads split evenly on valence but not on temperature. Roughly half of each thread coded ally, half opposition. Part Two ran proportionally hotter on the ally side: ALLY-H (ally-heated) was 28 percent of Part Two turns versus 12 percent of Called It. Same audience. Same argument. The Part Two piece drew heavier delivery from readers who agreed with it.

4. What surprised me

Three things I want to point out.

First, one exchange did what the essay was hoping someone would do. An opposition-adjacent commenter engaged in three sustained replies across the Called It thread. He delivered real substantive counters: the al-Awlaki 2011 legal framework, Article II and maritime authorities, the tax-loophole analogy for finding executive authority when Congress will not act. Partway through, he noticed the Substack byline listed a doctorate in Organizational Change and Leadership and wrote that he was “in for it.” He closed the exchange with a line about restoring his faith in humanity. One exchange out of 168. That fraction is what tells me the conversation is possible.

Second, even the substantive engagements stayed one register above the essay’s actual subject. The commenter above engaged al-Awlaki, Article II, maritime authorities, Shinseki, and JAG review as an ongoing process. He did not engage the JAG-office removals, the IG dismissals, the chaplaincy replacements, or § 1034 specifically. His argument sat at the culture-change / executive-workaround / majoritarian-mandate register. The essay’s argument sat at the oversight-channel level. The two registers did not meet, even in warmth. That gap is what tells me the frame shift is the hardest ask.

Third, extralegal-punishment proposals showed up on both sides. Some ally-side comments proposed the Secretary of Defense be routed straight to hell, Leavenworth, or Guantanamo. Some opposition-side comments proposed military tribunals for the six members of Congress who released the video. The grounds are not equivalent — impeachment of a Secretary of Defense is a real institutional path; congressional privilege protects the six from a tribunal proposal on its face — but the impulse to route around ordinary process shows up on both sides. I coded the ally-side extralegal proposals under D1 (loyalty-frame) because that was the closest fit. On next pass I will build a distinct code for them.

5. What I would do differently next time

Add a code for extralegal-punishment proposal. Big enough on both sides to deserve its own row rather than being folded into loyalty-frame.

Distinguish first-time commenters from repeat commenters. I did not track this. A first-time commenter’s delivery mode reads differently from a habitual commenter’s, and that distinction would sharpen the ally-side H2 finding.

Track thread velocity. A thread that draws 128 comments in the first six hours codes differently from one that draws 128 comments across four days. The essay’s finding might be sharper if I could show that architecture-silence held constant across both fast and slow coding windows.

Get a second coder. Solo coding is inter-reliably weak by definition. The next installment will run at least one code family past a second reader (Doug Krugman has offered) so the workbook carries a real inter-coder agreement number.

6. The repair

The Monday essay makes one analytical claim. The pattern belongs to the instrument, not to the individuals.

The instrument here is the political discourse machinery that trains readers to argue about persons (the Secretary, the President, the six members of Congress) rather than about mechanisms (the four oversight channels). The commenters, on both sides, did not choose to skip the architecture. The instrument selected against it. The workbook shows the selection at row level.

That is what makes the coded workbook a research artifact rather than a summary. A summary tells you what people argued about. A workbook lets you check whether the summary is right — and lets you take the same instrument to a thread of your own and see whether the pattern replicates.

Every subscriber who does this replicates the essay’s method. Enough replications and the analytical claim strengthens or falls. That is what I want from this series.

7. Notes on method

Anonymization. Every commenter’s name has been replaced with a stable respondent ID (R-001, R-002, etc.). The ID is consistent across the workbook, so exchanges and cross-thread patterns remain readable. Names appearing inside verbatim comment text have been redacted to [respondent]. Public officials named in comments (Hegseth, Mattis, Obama, Wilson, Holsey, Marshall, West) remain named. A private name-mapped master workbook stays on my working system for coding accuracy and audit; no subscriber ever receives the mapping.

Coding process. Two passes. First pass was open coding at the comment level, with codes generated inductively as new patterns appeared. Second pass was focused coding using the refined codebook, with every comment re-coded against the full scheme. Where a comment carried multiple substantive claims, it received multiple codes. Where a comment was pure delivery with no substantive counter (an insult without a claim), it coded H2 alone.

Reliability. Solo-coder installment. Inter-coder agreement is not reported for this installment. The next installment will include at least partial second-coder agreement on one code family.

Reproducibility. The raw thread capture ships with this installment. Any subscriber can re-code the same threads against the codebook and check where their coding matches mine or diverges. Divergences are interesting, not problematic. Send me one that surprises you, and I will publish it in the next installment.

Corrections made before publication. Two, and both are important because they show what the dual-workbook system is for.

First, the v1.0 codebook document described a different scheme than the one applied to the data. It named an M family that does not exist in the workbook and left twelve applied codes undefined. Worse, four shared code IDs carried different meanings in the two documents. Nothing in the Monday essay depended on the bad document, because the essay was written off the workbook. The codebook was regenerated from the workbook’s Code Book tab and shipped as v1.1.

Second, the Monday essay reported five subject counts that traced to no workbook row. Two of the five were code counts relabeled as subject counts; three had no source (me working alone at 1 AM). I coded all 168 turns for subject, published the boundary rules, added a Subject column to the workbook, and corrected the essay in place with a dated note on September 9, 2026. The largest change ran the other way from what a defensive writer would want: the Congressional video drew 30 comments, not the 14 I reported. More readers argued it than I said, and still none argued the architecture.

Neither error would have survived a subscriber opening the workbook. That is the argument for sharing the workbook.

Threats to validity. Selection bias — these threads sit on my page and were seen by my readers first. Timing bias — the September 3 post appeared during a specific news cycle that likely shaped what readers came in ready to argue. Coder bias — I am the author of the essay the workbook was built to support, which is the biggest single threat to the finding. The dual-workbook system, the reproducibility offer, and the second-coder plan for future installments are the mitigations. They are partial mitigations, not solutions.

This is Coded Threads №01, published September 9, 2026, free to read. The series runs monthly once the paid tier opens at five hundred subscribers. If you want it, pledge.

Jerry W. Washington, Ed.D., served 23 years in the Marine Corps. He holds a doctorate in Organizational Change and Leadership from USC Rossier and teaches at UC Irvine Division of Continuing Education. He publishes What Time Binds at what-time-binds.com.

Questions on this installment: info@what-time-binds.com

Subscribe now

Six months ago I wrote that the word ORDER was doing something dangerous inside the U.S. military. That the meaning regime behind it had shifted. That the standard channels for surfacing that shift (JAG, Inspector General, chaplaincy, the whistleblower protections in 10 U.S.C. § 1034) were designed for a meaning of “order” that no longer described what commanders were saying in briefings.

I promised Part Two.

Six months later, Part Two writes itself. The pattern I described from complaints has become documented practice. And the people who tried to reassert the constitutional meaning of “order” from outside the chain of command have been named as seditious by the President of the United States.

Six of them.

Share

The binary

The most common misconception about military service in American civic life is that a soldier who receives an order has two choices: comply or resign.

That misconception is wrong under U.S. law. It is wrong under U.S. military doctrine. It is wrong under the eleven general orders every Marine memorizes on the first day of boot camp.

It is also politically useful.

The binary shows up on both sides of the aisle. It appeared in a comment on my Facebook thread this week: “if you have concerns about the orders you’re being given you have but two choices, you either get behind and follow your orders to the best of your ability even if you don’t agree with them 100% or you resign if you can’t in good conscience follow them, there is no middle ground.” That commenter was writing in good faith. His statement is incorrect as a matter of law.

The same binary appears in every American conversation about military war crimes. My Lai. Abu Ghraib. The torture memos era. The Blackwater killings. The Kabul evacuation strike. The public defense of the soldiers who committed the acts almost always includes some version of: they had no choice, they were following orders. That is folk theory. It is not the law.

And it is the theory of civilian control that the current administration is banking on.

If the country believes soldiers have only two options, comply or resign, then a purge of officers who refuse to comply looks like normal turnover. A Chairman of the Joint Chiefs fired for asking questions looks like an executive prerogative. A Secretary of Defense telling four-stars they “will find the door” looks like accountability. A Letter of Censure to a retired Naval Captain for reminding troops of their legal duties looks like discipline.

The binary makes all of that legible. The binary makes the purge coherent.

Six members of Congress, all with military or intelligence backgrounds, released a ninety-second video in November of 2025 saying, in effect: the binary is not true. There is a third option. You are legally required to refuse unlawful orders. You have statutory protection when you report them. This is the doctrine. This is the oath. This is what your training says.

The President called it sedition and asked for their execution.

The video

The video was ninety seconds.

It was directed at active-duty service members and members of the intelligence community. The message was that they were not required to follow illegal orders. That they had legal protection when they refused. That the oath they took was to the Constitution, not to any individual.

It cited no new doctrine. It named no new law. It reminded uniformed personnel of what their basic training had already told them. Legal experts consulted by FactCheck.org confirmed the video did not meet any recognized definition of sedition under U.S. law.

That was the whole of it.

The response

On November 20, 2025, the President posted on his own platform, in capital letters: “SEDITIOUS BEHAVIOR, punishable by DEATH.” He shared posts calling for the six to be hanged.

The FBI opened an inquiry. The Department of Justice presented the case to a grand jury.

The grand jury did not indict.

In February 2026, Judge Richard J. Leon of the U.S. District Court for the District of Columbia granted Senator Mark Kelly a preliminary injunction against the Secretary of Defense, ruling that “Defendants have trampled on Senator Kelly’s First Amendment” protections.

The punishment did not stop there.

The Six

Senator Mark Kelly (D-AZ). Retired Navy Captain. F/A-18 combat pilot with four operational deployments. Former NASA astronaut with four spaceflight missions. Signatory of the video. In January 2026, Secretary Pete Hegseth issued Kelly a formal Letter of Censure and initiated retirement grade determination proceedings, a process that can reduce the rank at which a retired officer is pensioned. Kelly sued. Judge Leon issued the injunction the following month.

Senator Elissa Slotkin (D-MI). Former CIA analyst, three tours in Iraq. Signatory. Reported receiving violent threats after the video was released. Named in the FBI inquiry.

Representative Chrissy Houlahan (D-PA). Air Force Captain, engineer. Signatory. Named in the FBI inquiry.

Representative Jason Crow (D-CO). Army Ranger, three combat deployments in Iraq and Afghanistan, Bronze Star. Signatory. Named in the FBI inquiry.

Representative Chris Deluzio (D-PA). Navy veteran, Iraq. Signatory. Named in the FBI inquiry.

Representative Maggie Goodlander (D-NH). Navy Reserve intelligence officer. Signatory. Named in the FBI inquiry.

Six people. Two senators. Four representatives. Twenty-plus combined years of active-duty and intelligence service. One ninety-second video reminding uniformed personnel of doctrine the Corps teaches on the first day of boot camp.

The four channels

Part One named four channels through which the meaning-regime shift I described should have been visible from inside the institution: the Judge Advocate General’s Corps, the Inspector General system, the military chaplaincy, and the Military Whistleblower Protection Act codified at 10 U.S.C. § 1034.

Between March and September of 2026, each of them was worked on.

JAG. In early 2025, the administration ordered the removal of the top JAG officers of the Army, Navy, and Air Force. The stated reason was that the incumbents might obstruct the President’s policy agenda. The JAG Corps is the internal legal review function of the military. Its purpose is to advise commanders on whether orders are lawful. Removing its senior leadership does not eliminate the legal review. It changes who reviews. (I wrote about the earlier stages of this pattern in Star Wars or Star Creep? The Battle Over America’s Generals, published on Medium in May 2025.)

Inspector General. In January 2025, seventeen federal inspectors general were dismissed in a single evening. The IG system exists to receive complaints of misconduct, including complaints from service members. Removing its senior leadership does not eliminate the complaint channel. It changes who receives complaints and who decides which are pursued.

Chaplaincy. In May 2025, monthly Christian worship services began at the Pentagon during working hours, livestreamed on the Department of Defense’s internal television network. In February 2026, Secretary Hegseth invited Doug Wilson, a theologian who openly advocates making the United States a Christian theocracy, to preach at the Pentagon. In December 2025, the Army Spiritual Fitness Guide was eliminated on the stated ground that it did not focus sufficiently on God. The chaplaincy exists to serve the spiritual needs of a religiously diverse force. Remaking its posture does not eliminate the chaplaincy. It changes what spiritual authority the institution is willing to name. (Part One documented the earlier moves in this sequence.)

10 U.S.C. § 1034. The Military Whistleblower Protection Act creates a statutory right for service members to raise concerns to the Inspector General, to a member of Congress, or to superior officers. Retaliation for a protected communication is itself unlawful. On December 12, 2025, Adm. Alvin Holsey handed over command of Southern Command after one year in a role typically held for three. Reuters reported that Hegseth had “grown frustrated” with SOUTHCOM over the Caribbean boat strikes. Hegseth’s spokesperson denied tension. Holsey himself said the decision was personal. Both Republican and Democratic lawmakers had publicly questioned the legality of the strikes. The whistleblower channel exists to receive exactly the kind of concerns those strikes have raised. What the record shows is that the admiral overseeing the operation left his command two years early. What it does not yet show is anyone who used the channel and was rewarded for using it.

What the channels did

The channels still exist. Their statutes were not repealed. Their offices were not closed. Their doors are still labeled.

What changed is that each channel was staffed with people whose willingness to use the channel against the person who appointed them is now the operative question. Part One’s argument was that the meaning regime behind the word ORDER had shifted. The corollary is now visible: the meaning regime behind INSPECTOR, ADVOCATE, CHAPLAIN, and WHISTLEBLOWER has shifted too. The forms remain. The referents have moved.

That is what the Six were trying to name from outside.

The eleven

If you’ve read my work for a while, you already know the background. If you’re new here — twenty-three years in the Marine Corps. General Order 3 was the first thing I memorized. It’s still the cleanest statement of what I mean when I write about the duty.

The Eleven General Orders of a Sentry are recited from memory by every Marine and Sailor from the moment they enter recruit training. Two of them answer the comply-or-resign question directly.

General Order 3: To report all violations of orders I am instructed to enforce.

General Order 9: To call the corporal of the guard in any case not covered by instructions.

Reporting is a duty. Escalating is a duty. They are not options. A Marine who sees an order being violated, including an unlawful order being given, is required to report it. A Marine who encounters a situation not covered by their instructions is required to escalate to higher judgment. That is Day One doctrine.

The comply-or-resign binary erases Day One doctrine and replaces it with folk theory.

The word “lawful”

Above the enlisted floor, the Uniform Code of Military Justice puts the same duty in statute.

UCMJ Article 90 makes it an offense to willfully disobey a lawful order.

UCMJ Article 92 makes it an offense to fail to obey a lawful order or regulation.

The word “lawful” is doing all the work.

An unlawful order is not a lawful one. An officer who follows an unlawful order can be prosecuted for it. That is not a modern innovation. It was codified into U.S. military doctrine after 1945, when the United States helped write the Nuremberg Principles that established personal legal responsibility for war crimes committed under orders.

The counter-case is My Lai. In 1968, an Army platoon commanded by Second Lieutenant William Calley massacred somewhere between 347 and 504 unarmed Vietnamese civilians, most of them women, children, and elderly, over a period of hours. Calley’s defense was that he had been following orders. The Army court-martial rejected the defense. Calley was convicted. The lawful-order distinction was the reason the conviction could hold.

Every officer in the U.S. military knows this. It is taught. It is examined. It is memorialized in institutional memory. The distinction between a lawful and an unlawful order is not a technical footnote. It is the moral core of the officer’s oath.

The Six were restating that core. (I wrote about a related failure of the professional ethic in the June essay What Do You Mean by Warrior Ethos?, which named General Chris Donahue by name a month before the Times printed the same pattern.)

Where things stand

Kelly’s Letter of Censure and his retirement grade determination proceedings remain in play as of this writing, though Judge Leon’s February injunction has slowed the process. The other five continue in their offices in Congress. The FBI inquiry did not produce indictments. The grand jury declined to bring charges. No court has ruled the video seditious. No court has ruled it unlawful. It was neither.

That is where things stand as the second-term Hegseth Defense Department continues to remove flag and general officers from promotion lists, most recently four Army one-stars in late August, part of a total pattern that the New York Times counted at forty-five officers pulled from promotion lists by August 2026, more than half of whom are female or Black.

The channels named in Part One did not surface any of it in time to prevent it. The Six did. The response was legal harassment, administrative discipline, criminal referral, and public calls from the President of the United States for their deaths.

Brian’s question, answered

At the end of Part One I put a question in Brian’s mouth. If a minister on Facebook could see the mechanism clearly in a single scroll, what would it take for the institution to see it from the inside?

The answer as of September 2026: it took six sitting members of Congress with military and intelligence backgrounds. And the response to their seeing it was a criminal referral, a Letter of Censure, and an official call for their execution.

The institution is not seeing it. It is being made to look away.

Brian said, back in March, that this was “ridiculously dangerous.” He meant it as diagnosis. It reads now as description.

Part Three, if it comes, will be about what happens when the outside institution — the Congress that oversees the military, the courts that review the discipline, the press that reports the removals — is asked to hold the line the internal channels no longer can.

The Six did their part.

The question is who does the next part.

Sources and further reading

Prior essays in this thread of argument

• When Order Becomes Oracle, Part One: The Word That Runs Everything (What Time Binds, March 5, 2026)

• What Do You Mean by Warrior Ethos? (What Time Binds, June 24, 2026)

• Star Wars or Star Creep? The Battle Over America’s Generals (Medium, May 6, 2025)

On the November 2025 video and its aftermath

• November 2025 United States military video controversy — Wikipedia overview and citations

• Trump says Democrats’ video message to military is ‘seditious behavior’ punishable by death — PBS NewsHour

• Experts Say Democratic Video Not ‘Seditious,’ as Trump Claims — FactCheck.org

• FBI opens inquiry into 6 Democrats appearing in video that urged defying illegal orders — CBS News

• FBI opens probe into Sen. Mark Kelly and 5 others who urged military to disobey illegal orders — Cronkite News

• Who is Mark Kelly, and why is the US senator accused of ‘sedition’? — Al Jazeera

On Adm. Alvin Holsey and SOUTHCOM

• Admiral hands over leadership of US Southern Command amid boat strikes and buildup in the region — Stars and Stripes, December 12, 2025

• Admiral Overseeing Caribbean Mission to Retire after One Year in Role — USNI News, October 16, 2025

• Top admiral’s retirement sets off alarm bells over Caribbean boat strikes — The Hill

On the Hegseth Defense Department and the promotion purge

• Hegseth blocks promotions of 4 Army officers with ties to sacked generals — The Hill, August 28, 2026 (companion coverage of the New York Times article of the same title)

Statutory and doctrinal references

• 10 U.S.C. § 1034 — Communicating with Members of Congress and Inspector General; prohibition of retaliatory personnel actions — Cornell Legal Information Institute

• 10 U.S.C. § 890 (UCMJ Article 90) — Assaulting or willfully disobeying a superior commissioned officer — Cornell LII

• 10 U.S.C. § 892 (UCMJ Article 92) — Failure to obey order or regulation — Cornell LII

• Nuremberg Principles — Wikipedia summary of the U.N. International Law Commission’s codification

Historical cases

• My Lai massacre

• William Calley

Part One of this series, “When Order Becomes Oracle: The Word That Runs Everything,” was published on March 5, 2026. Brian Cobb is quoted with his knowledge and permission. He holds a degree in biblical studies and has served in pastoral ministry for decades.

If this piece was useful to you, share it with someone who works in high-stakes environments. This is what meaning repair looks like when the stakes are real.

Subscribe now

November 2024. Five of us in a car. My wife, my three daughters, and I, all wearing I Voted stickers, my youngest with hers on her nose. She was twenty. It was the first November my whole family voted together.

I have voted in every federal election of my adult life. For twenty-three years of that record I was in uniform, mailing a Federal Post Card Application to a Kansas county clerk from wherever I was stationed. Camp Pendleton. Austin. Okinawa for six years with my family. Camp Leatherneck in Helmand. Home of record: Kansas. The envelope always went to the same county. The Uniformed and Overseas Citizens Absentee Voting Act had been the ritual of my adult life. The Military and Overseas Voter Empowerment Act, signed on October 28, 2009, as part of the FY2010 National Defense Authorization Act, streamlined the machinery. I trusted the machinery because the machinery worked.

This piece is about what happened between November 2024 and this August. What has been done around voting in the last twenty-two months.

November 2024. First November my whole family voted together. My youngest had just turned of age.

Share

What is settled

On March 31, 2026, President Trump signed Executive Order 14399, “Ensuring Citizenship Verification and Integrity in Federal Elections.” 91 Fed. Reg. 17125. The Order directs the Secretary of Homeland Security to compile Citizenship Lists and send them to each state (§2(a)). It directs the Attorney General to “prioritize the investigation and, as appropriate, the prosecution” of state and local election officials who issue ballots to ineligible voters (§2(b)). It directs the Postal Service to initiate rulemaking to require unique barcodes on every outbound ballot envelope and to allow USPS to refuse to transmit ballots from any voter not enrolled on the appropriate list (§3).

Twenty-three states and the District of Columbia sued in the District of Massachusetts. The district court enjoined §§2 and 3. The First Circuit refused to stay the injunction, 2-1. On Monday, August 24, 2026, the Supreme Court granted the stay, 6-3. The Trump v. California per curiam is unsigned. Justices Sotomayor and Kagan dissented. Justice Ketanji Brown Jackson dissented separately, alone, in twenty-three pages, and called the ruling “a grave misuse of our equitable emergency powers.”

The Department of Justice’s Voting Section, the office that would enforce federal voting-rights law against the state and local officials the Order threatens with prosecution, has three attorneys left.

One Justice kept a ledger

Justice Ketanji Brown Jackson joined the Court in June 2022. Since then, she has dissented in four voting-rights and election-law rulings where the majority sided against voters, minority communities, or state election officials. The dissents are on the record. Read them in order, and a pattern shows up.

May 23, 2024. Alexander v. South Carolina State Conference of the NAACP. 6-3 on racial gerrymandering. Jackson joined Justice Kagan’s dissent, which called the district court’s factual findings that South Carolina drew a race-based congressional map “more than plausible.” The majority set those findings aside.

August 22, 2024. Republican National Committee v. Mi Familia Vota. Emergency stay concerning Arizona’s proof-of-citizenship voter registration law. Jackson dissented from the partial grant of the stay. Four Justices, Jackson among them, would have denied the RNC’s application entirely. The case was about a state requiring documentary proof of citizenship for voter registration. Executive Order 14399 takes the same premise national.

April 29, 2026. Louisiana v. Callais. 6-3 on Section 2 of the Voting Rights Act. The Court held that Section 2 does not require Louisiana to keep two majority-Black congressional districts despite African Americans making up more than 32% of the state’s population. A week later, on May 4, 2026, when the Court accelerated the ruling’s implementation, Jackson wrote a four-page dissent to the follow-up order and called the Court’s action “unwarranted and unwise.”

August 24, 2026. Trump v. California. Jackson wrote alone in twenty-three pages, called the ruling “a grave misuse of our equitable emergency powers,” and closed with the line: “This is not a close call.”

Alongside those dissents, Jackson has spoken publicly about the Court’s use of the emergency docket. In an April 2026 speech at Yale Law School, she said the practice has a “corrosive effect” on the judicial system.

The pattern she has been documenting from inside the institution is the pattern this essay describes from outside it. The institution has been inconsistent. She has not.

What is being pinned right now

The Court’s ruling did not decide the lawfulness of the Executive Order. It decided that the states cannot be in court yet. That distinction matters because the Order is being implemented while you read this. USPS promulgated its proposed rule on June 2, 2026. The Government blew past the July 29 deadline the Order set for a final rule, then issued that final rule on August 21, 2026 — three days before the Supreme Court's ruling and while the District Court's injunction against implementing §3 was still in force. The Federal Register formally published it today, August 26. Justice Jackson names the timing choice as strategy: the executive controls when a rule is proposed and finalized, and delaying gives it a court order that blesses the strategy of proposing contested changes, prompting chaos, and keeping valid legal claims at bay.

Here are the words being pinned this week. Read them as a sentence. Together they answer who gets to vote in November.

INTEGRITY

Plain definition. The property of a process that produces accurate results without corruption. In election law, it names chain of custody, signature verification, and audit trails.

What “integrity” is doing. It is the frame that lets a President reorganize the machinery of state elections by executive order, and lets the Republican National Committee sue overseas voters in swing states without evidence. In October 2024, Judge Sima Patel in Michigan and Judge John Smith in North Carolina both dismissed RNC lawsuits challenging overseas ballots, in each case finding no evidence of fraud. The lawsuits used “integrity” as their basis. The evidence side of the word was empty.

Pin. Election integrity is a property of a process, verified by audit and observation. It is not a permission slip for anyone with power to reshape who counts. Includes: chain of custody, signature verification, post-election audit. Excludes: prosecuting state officials for issuing ballots the Attorney General does not want issued. Example: a bipartisan post-election audit. Non-example: EO 14399.

MONITORING

Plain definition. Watching a process to detect and correct deviations. In federal voting-rights law, it names the Justice Department’s Voting Section observing polling places under the Voting Rights Act.

What “monitoring” is doing this week. The DOJ still uses the word. The office behind the word has three attorneys.

Pin. Monitoring is enforcement staffed by people with authority to act. A word used to describe an empty office is not monitoring. Includes: federal observers with legal authority under §2 of the Voting Rights Act. Excludes: the chilling of state and local election officials by federal prosecutors. Example: DOJ election observers in the 1965 Selma federal registration effort. Non-example: three attorneys.

ELECTION INTEGRITY ARMY

Plain definition. None. The phrase is a brand.

What “election integrity army” is doing this week. It labels four operations under one phrase. Poll watching by trained partisan observers, which is legal in most states under state rules. Ballot chasing, which is a term for coordinated absentee-ballot collection and is regulated state by state. Voter intimidation at drop boxes, which federal law prohibits. Private surveillance of election workers, which crosses state harassment lines and, at some scales, federal ones. One phrase, four operations, only one of which the private label makes legal.

Pin. “Army” is a legal category. A self-appointed civilian group is not one. Private groups that describe themselves as an army performing election duties are performing four different acts, at least one of which is a federal crime. Includes: state-credentialed poll observers with rules and complaint procedures. Excludes: armed private groups at drop boxes. Example: certified election observers.
Non-example: masked men in tactical vests at a drop box.

INTERNAL DIRECTIVE

Plain definition. An order from a President to his own subordinates that binds no one outside the executive branch. It usually names scheduling memos, coordination directives, budget guidance.

What “internal directive” is doing this week. It is the frame the Trump v. California majority used to say Executive Order 14399 is not reviewable in court because the Order addresses federal officials. Justice Jackson’s reductio is the pin: imagine a “purely internal” directive from a President to the Postal Service ordering it not to deliver mail-in ballots in disfavored states. That would still be a directive to federal subordinates. It would still be unreviewable under the majority’s frame. The label does not survive its own logic.

Pin. An order is “internal” if it does not require action by anyone outside the executive branch and does not create the threat of federal action against anyone outside it. EO 14399 requires state cooperation and threatens state officials with prosecution. It is not internal. Includes: scheduling and coordination inside the executive branch. Excludes: an order that runs on state cooperation and is enforced by federal prosecution of state officials. Example: a memo setting the executive’s work-from-home policy. Non-example: EO 14399.

STANDING

Plain definition. The legal doctrine that decides who is allowed to sue. It exists to keep courts from hearing hypothetical cases.

What “standing” is doing this week. It is the doctrine the Court used to close the courthouse door on twenty-three states with constitutional duties to administer elections. Under Trump v. California, states that are already spending money, retraining officials, and rewriting guidance because of the EO have no concrete injury. Under Purcell v. Gonzalez (2006), the same states cannot file after the final rule issues because that filing would come too close to the election. Too early now. Too late later. Justice Jackson names the trap in a footnote.

The Court decided a related case, Bost v. Illinois State Board of Elections, less than a year ago. In Bost, the Court held that a political candidate had standing to challenge an election rule change based on a general fair-election interest. Today the Court holds that twenty-three sovereign states with constitutional duties to administer elections do not. Jackson: “This means either Bost is wrong or today’s decision is indefensible (or both).”

Jackson herself dissented in Bost. She thought the candidate should not have received standing either. Her point in Trump v. California is that the Court cannot grant standing there and deny it here in the same twelve months without contradicting itself.

Pin. Standing exists to prevent hypothetical cases from being heard. It has been used this week to prevent a documented, imminent case from being heard because the executive can delay its own finalization. That is the opposite of what the doctrine exists to do. Includes: gatekeeping for cases with no real injury. Excludes: gatekeeping for cases whose plaintiffs are already suffering the injury the doctrine was written to require. Example: dismissing a suit over an executive order that has been rescinded. Non-example: dismissing a suit by state election officials who have already reallocated staff and rewritten guidance because of the order.

What I am not saying

I am not saying every executive order is illegal. Presidents issue orders. Some are lawful. Some are not. Some are unreviewable for reasons that make institutional sense. That is not the piece.

I am not saying every 2026 vote will be blocked. Millions will be cast and counted. Kansas will send ballots to its overseas voters starting forty-five days before November 3. My daughters will vote again. That is not the piece either.

I am not saying the Supreme Court is corrupt. Corruption is a specific claim requiring specific evidence. The claim I am making is different. I am pointing to doctrine used against its purpose. Standing exists to prevent hypothetical cases. It was used this week to prevent a documented case. That is a doctrinal failure, and it is bad enough on its own.

I am not saying Republicans invented voter suppression. The record on both parties is long. The claim I am making is about a specific week in which a President, a Court, and a private label all pinned the same five words in the same direction, and about what those five pins do to a November election that is seventy days away.

I am not saying citizenship should not be verified. Every state already verifies eligibility at registration. The claim I am making is about what happens when a federal Citizenship List (one the Government has conceded will be incomplete) becomes the gate through which state ballots must pass, and when state officials who mail ballots to voters not on that federal list face federal prosecution.

What I am saying

The words being pinned this week decide who gets counted before anyone casts a ballot in November. Turnout is the last step. The words are the first step, and the first step is happening now.

I am writing as a UOCAVA voter who trusted a federal machinery for twenty-three years and cast a ballot alongside his wife and daughters in 2024. Every word I have pinned in this piece touches that vote. INTEGRITY is the frame the RNC used in 2024 to try to invalidate ballots like the ones I mailed. MONITORING is what the DOJ office with three attorneys still calls what it no longer has the staff to do. ELECTION INTEGRITY ARMY is what private groups call themselves when they replace the office with three attorneys. INTERNAL DIRECTIVE is what the Court called an executive order that reaches into every state’s ballot-mailing system. STANDING is what the Court used to keep the states that administer those systems out of court.

The five pins are a sentence. The sentence says: the executive branch and the Court have arranged the words so that the challenge to the executive branch cannot be heard in time to matter.

One Justice on that Court has been documenting the arrangement since 2024. Her ledger is public.

What the reader owes

Read the words before you accept them. When someone says integrity, ask who wrote the list. When someone says monitoring, ask how many attorneys. When someone says army, ask which of the four operations they mean. When someone says internal, ask who is being prosecuted. When someone says standing, ask who is being kept out.

Register if you have not. Read your county’s rules for absentee and mail-in voting. Read your ballot before you fill it out. Understanding the ballot is the first thing an accountable public owes itself.

Read the dissent, too. Someone on the inside has been reading the institution for you.

November 2024. Five of us in a car. My youngest with the sticker on her nose. It was the first November my whole family voted together.

Twenty-two months later, the Court told twenty-three states they came too early. Purcell will later tell them they came too late. One Justice on that Court has been keeping the ledger since 2024. This week she wrote twenty-three pages naming the trap and called the ruling a grave misuse of our equitable emergency powers.

The last remaining check on definitional control is a reader who moves slowly, checks the source, and refuses to accept the word as given.

Read the words. Read the ballot. Read the institution. Read the dissent.

Note, August 26, 2026 (afternoon). This piece originally described the Government as having blown past the July 29 deadline for a final USPS rule — accurate as of Justice Jackson's dissent. USPS issued that final rule on August 21, 2026, three days before the Supreme Court's ruling and while the District Court's injunction against implementing §3 was still in force. The Federal Register formally published it today, August 26 (Doc. 2026-17238). The clause in "What is being pinned right now" has been sharpened to reflect that timeline. Democratic plaintiffs (DSCC, DCCC, DNC, Democratic Governors Association, Schumer, Jeffries) filed a supplemental complaint and renewed motion for preliminary injunction against the USPS rule on August 25, 2026 in the U.S. District Court for D.C. Hearing sought on or around September 4.

Subscribe now

What Time Binds studies how meaning drifts and how we repair it. One field guide like this every week, free.

How I make this: reporting and verification are mine, drafting is a collaboration with AI tools I review, and corrections get pinned, never buried. Full process: what-time-binds.com/how-i-make-this

Subscribe now

Monitoring

Yesterday’s Department of Justice press release, number 26-939, used the word throughout. Election monitoring. Monitoring polling sites. Monitoring compliance with federal civil rights laws. Two Civil Rights Division attorneys in Laramie County. Four in Miami-Dade. Assistant Attorney General Harmeet Dhillon called it “an ongoing priority for this office” and said, “nondiscriminatory monitoring ensures all elections remain free, fair, and accessible to all.”

Same word 1965 used. Different office, different docket, different mission memorandum. The word did not carry the same weight in any two places it appeared.

Three meanings under one word

Three definitions reside under “monitoring.” Each carries the weight of the others. Each speaker acts as if their definition is the only one on the table.

Meaning 1: Compliance observation. The 1965 meaning. A federal attorney at a polling site checking that a covered jurisdiction is not violating the Voting Rights Act. Section 2 on discriminatory effects. Section 203 on language assistance. The National Voter Registration Act on maintenance procedures. The observer is trained to a published standard. The office has a docket. The finding gets filed and becomes evidence in a case.

Meaning 2: Federal deterrence presence. Federal officers at a polling place in a country whose president has repeatedly called elections illegitimate. The voter reads the badge and the seal, and calibrates. Nothing is said. Something has changed. The word for this in enforcement literature is “chilling effect.” It does not require action to work. Presence is the action.

Meaning 3: Records audit. The administrator’s meaning. Checking that a jurisdiction followed procedure. Polls opened on time. Provisional ballots logged. Assistance offered where required. This is what Dhillon’s phrase “monitoring compliance with federal civil rights laws” sounds like. It sounds procedural. It sounds neutral. It sounds like meaning 1 without the docket behind it.

Sitting under all three, the actual referent for the 1965 word: monitoring as a Voting Rights Act enforcement function conducted by the Civil Rights Division’s Voting Section, with statutory authority, published mission, and career-attorney practice.

What happens when they collide

When one press release uses the same word for all three meanings, the reader who trusts meaning 1 assumes career VRA attorneys are watching for suppression. The reader who fears meaning 2 sees federal officers at the polling site of a county the administration lost by twenty points. Both are looking at the same word doing different jobs.

My July 30 essay How Exactly? named the mechanism at scale. The Justice Department’s Civil Rights Division lost roughly 70 percent of its attorneys between January and May 2025. The Equal Employment Opportunity Commission lost its quorum in January 2025 and deprioritized disparate-impact claims. In April 2026, Louisiana v. Callais shifted Section 2 of the Voting Rights Act from an effects test to an intent test. The federal workforce contracted by 271,825 employees between September 2024 and January 2026, with Black women losing 95,371 positions.

The words “civil rights,” “compliance,” and “monitoring” survived every one of those actions. What each word protected changed underneath.

Share

Federal vote monitoring, 1870–2026: same word, four eras

1870–1894. Federal monitoring created, then stripped. Congress passed the Enforcement Act of 1870 to make the Fifteenth Amendment enforceable. President Grant signed it on May 31, 1870. Section 9 gave federal marshals arrest authority in election cases. Section 13 authorized the president to use the militia and army. The Second Enforcement Act of 1871 permitted federal supervision of any election in a town over twenty thousand people when just two citizens requested it. Federal monitors were protection agents. The Supreme Court then gutted the criminal provisions in United States v. Cruikshank and United States v. Reese (1876). Democrats won the presidency and both houses of Congress in 1892. Two years later, Congress passed and President Cleveland signed the Civil Rights Repeal Act of 1894 (28 Stat. 36), eliminating the federal-supervisor structure Congress had created twenty-four years earlier.

1894–1965. The word without the office. Mississippi ratified its 1890 constitution with literacy tests, poll taxes, and “understanding” clauses. Black registration in Mississippi fell from a Reconstruction-era majority to roughly six percent of the eligible Black population within a decade. By 1964, 6.7 percent of eligible Black Mississippians were registered. In Sunflower County, fewer than 200 of 13,000 eligible Black adults. In Leflore County, roughly 250 of 30,000. Federal monitoring did not disappear from public speech. The word migrated. Local “poll watchers,” sheriffs, and party officials now did the watching. Their subject was Black voters attempting to register. Presence was the enforcement. The word “monitor” continued to describe the mechanism. Whose enforcement had changed.

1965–2013. Federal monitoring restored. The Voting Rights Act of 1965 revived the federal monitoring apparatus. Sections 6 and 8 authorized federal examiners and observers in covered jurisdictions under the Section 4(b) coverage formula. Career Voting Section attorneys carried Section 2 discriminatory-effects cases. Mississippi Black registration rose from 6.7 percent in 1964 to 59.8 percent by 1967. Two years. That is what monitoring did when the word carried authority.

2013–present. The word after the authority. Shelby County v. Holder, 570 U.S. 529 (2013), struck down the Section 4(b) coverage formula, effectively deactivating federal observer deployment under Section 8. Deployments fell from 51 jurisdictions in 23 states in 2012 to 28 jurisdictions in 18 states in 2014. A 45 percent decline in two years. Section 2 authority continued. Career Voting Section attorneys carried the work through 2024. In 2025, the Voting Section was cut from roughly thirty career attorneys to three. Section 2 cases in Georgia, Texas, and Louisiana were withdrawn. In August 2026, the Justice Department announced monitor deployments to Miami-Dade County and Laramie County. The mission memorandum authorizing that work has not been produced.

The correlation to the Jim Crow decades is not metaphorical. It is documented. The word “monitoring” survives federal withdrawal. What it does under the same word depends on who is doing the watching, and against what standard.

What monitoring actually means

Monitoring is watching a documented standard with the authority to act when it is violated. Without the standard, without personnel trained to it, without a published trigger for action, the word collapses into presence.

That definition is testable. It survives contact with the Voting Rights Act. It survives contact with EEOC guidance. It survives contact with the Federal Election Commission’s own compliance manuals. It does not survive contact with a press release that keeps the word and drops everything the word once required.

Includes/Excludes

Includes:

• A Civil Rights Division attorney observing a jurisdiction under a published mission memorandum tied to Sections 2, 203, or 208 of the Voting Rights Act

• Post-observation report to the jurisdiction with a documented finding on the record

• Section 2 litigation with a filed complaint and a public docket

• Language-assistance compliance under Section 203 with published criteria for jurisdictional coverage

• A Voting Section deployment with attorneys trained to the current mission memorandum

Excludes:

• Federal presence at a polling site with no published criteria for site selection

• Deployment in a jurisdiction where the enforcing office holds no attorneys trained to the current mission

• Enforcement language tied to no specific case, complaint, or docket

• A mission memorandum withheld from Senate oversight while the enforcement it authorizes is publicized

Example. The 2022 Justice Department midterm deployment. Twenty-four states. Sixty-four jurisdictions. Announced weeks in advance with published criteria. The Voting Section then had roughly thirty career attorneys with an active docket of Section 2 cases in Georgia, Texas, and Louisiana. That is monitoring.

Non-example. The August 18, 2026 deployment. Two counties. Six attorneys. Announced the day of the primary. The Voting Section at a watchdog-estimated three attorneys. The mission memorandum requested by Senator Alex Padilla on May 6, 2026 and not produced. The Section 2 cases in Georgia, Texas, and Louisiana withdrawn. That is presence.

The three questions

When an institution announces it is “monitoring” something, take about a minute and ask three questions.

1. “What documented standard is the monitor watching for? Show me the mission memorandum, the statute section, or the compliance manual.”

2. “Who was trained to the standard? Where is the roster?”

3. “What action happens when the standard is violated? Show me the last case.”

If any of the three go unanswered, the word “monitoring” is doing work its practice does not support. File the silence next to the announcement.

File this next time

Before you accept or reject an institution’s claim that it is monitoring something, log two things.

1. Which of the three meanings am I being invited to hear? Compliance observation, deterrence presence, or records audit.

2. Which published standard is the institution watching for?

If you cannot answer the second question, you are not watching monitoring. You are watching presence. Presence is a policy choice. Presence is not neutral. Presence is what the observer wants the observed to see.

Name the standard, or name the silence.

This is a Field Guide entry. Companion to the July 30 special essay, How Exactly?, which named the mechanism at scale.

Related WDYM entries: What Do You Mean by Election Integrity?, where the same word was doing similar work. WDYM: DEI, where three definitions in one word ran the same failure mode at a different scale.

Subscribe now

Sources

Current record:

• Department of Justice Office of Public Affairs, Press Release 26-939, “Justice Department to Conduct Election Monitoring in Florida and Wyoming Primary Elections,” August 18, 2026.

• Democracy Docket, “DOJ Voting Section Has Just Three Lawyers Left, Watchdog Estimates” (attorney count estimate).

• Senator Alex Padilla, letter to Assistant Attorney General Harmeet Dhillon, May 6, 2026.

• CBS News reporting on Bloomberg interview with AAG Dhillon regarding 2026 midterm monitor deployment, August 17, 2026.

• Axios, “DOJ to monitor polls in 24 states for compliance with voting rights laws,” November 7, 2022.

• What Time Binds, “How Exactly?” July 30, 2026.

Statutes and cases:

• Enforcement Act of 1870, 16 Stat. 140 (May 31, 1870); Second Enforcement Act of 1871, 16 Stat. 433.

• United States v. Reese, 92 U.S. 214 (1876); United States v. Cruikshank, 92 U.S. 542 (1876).

• Civil Rights Repeal Act of 1894, 28 Stat. 36.

• Voting Rights Act of 1965, Public Law 89-110. Section 2 codified at 52 U.S.C. § 10301; Sections 6 and 8 codified at 52 U.S.C. §§ 10305, 10310.

• Shelby County v. Holder, 570 U.S. 529 (2013).

Historical record:

• U.S. Commission on Civil Rights, The Mississippi Delta Report (Mississippi Black voter registration 1867, 1890, 1954-55, 1964, 1967).

• Campaign Legal Center, “Failure to Observe: The Federal Observer Program After Shelby County” (2012 vs. 2014 deployment counts).

On July 20, 2026, the Republican polling firm co/efficient sent a two-page memo to House Republican Conference member offices. The header marks it CONFIDENTIAL. The footer reads “Prepared by co/efficient for the House Republican Conference.” Ryan Munce, the firm’s president, signed it. Washington Examiner reporter Hailey Bullis obtained a copy, and it has since surfaced online.

Page 1 carries a section heading: The Only Variable Is the Words.

Under that heading, Munce compares two survey questions that test the same tax policy. Framed in the third person, about growth and the national debt, the conservative position wins by 14 points and loses moderates 51 to 37. Framed in the second person, about whether workers should keep more of their own money, the same position wins by 54 and carries moderates 65 to 18.

Then Munce writes the sentence:

“The gap between +14 and +54 is not a different country; it is a different sentence.”

He is right. He measured what a sentence is worth, priced it at 40 points of margin, and sold the finding to House member offices.

Page 2 applies the same finding to three words that name three different things.

The word

“Communist.”

Why it is a magnet

The memo asked 2,634 registered voters a single question: “What do you consider the most dangerous system of government for the United States to pursue?”

Communism, 42 percent. Socialism, 15. Marxism, 6. None of the above, 24. Unsure, 12.

Munce’s reading of his own chart: “Americans still hold communism as far more dangerous and alarming than socialism and Marxism. That is an important indicator for Members: do your best to connect all three together, because these government systems are fruit of the same poisonous tree.”

The guidance section says it plainly. Bullet three:

“Call them what they are: Socialists, Marxists, Communists. Communism rings the loudest alarm with constituents (42 percent name it the most dangerous system), so do not retreat to the softer label just because socialism alarms fewer in the abstract. Connect all three, fruit of the same poisonous tree, and let the family name carry the alarm.”

Read that last clause again. Let the family name carry the alarm. The memo concedes the three words are not the same. Its own polling separates them by 27 points and 36 points. The instruction is to merge them anyway, so the fear attached to the most feared word travels to the other two.

The memo never argues that any Democrat is a communist. It does not make the claim once. It recommends the word for how the word polls.

Since then, the word has done exactly what a memo like that intends. On July 3 at Mount Rushmore, President Trump called the “communist menace” a “mortal threat.” On July 22 in Georgia: “We have some great candidates running, they got to win by a lot; otherwise we’re going communist.” In a July 28 endorsement video for Wisconsin candidate Michael Alfonso, the “radical left communists in Congress.” Speaker Mike Johnson has framed the midterms as a choice about whether the country goes “down this dark road of death to communism.” House Republican Conference Chair Lisa McClain: “the failed nightmare of communism.”

One word, moving fast, covering at least six different things and my Facebook thread.

The failure mode when meanings collide

Abdul El-Sayed won Michigan’s Democratic Senate primary on August 4. He rejects the democratic socialist label. Trump called him a communist the day before the vote.

Darializa Avila Chevalier beat thirteen-term Representative Adriano Espaillat in a June primary in New York. She once posted “seize the means of production” from an account she has since deleted. She told MS NOW she is not a communist and called the focus on the old posts a distraction.

Two candidates. Opposite facts. The same word applied to both.

Now put a voter in front of that. A Michigan voter hears “communist” on Monday about a candidate who has declined the socialist label. A New York voter hears “communist” about a candidate whose old posts quote Marx. The word did not change. The thing it points at changed completely, and nothing in the sentence tells the listener which one arrived.

That is what a magnet phrase does. It absorbs incompatible cases under one banner and offers no way to sort them. The voter cannot check the claim, because the claim has no fixed referent to check.

The memo built this on purpose. Bullet four instructs members to “take the same words to every district,” with “one message for the whole country; no separate playbook required.” A word that means six things is the only kind of word that can do that job.

The pin

“Communist,” in the 2026 midterm campaign, is an umbrella term deployed by design. A July 20 memo from co/efficient to House Republican Conference member offices instructs members to merge socialist, Marxist, and communist so the alarm attached to the most feared word transfers to the other two. The word currently covers at least six distinct positions with different programs and different histories.

When someone uses it, ask which one they mean, and ask what they want done about it.

Boundary test

What the words actually name:

Communism, as a system of government, means a single-party state with state ownership of the means of production. That is what 42 percent of respondents ranked most dangerous.

Marxism is a body of economic and historical analysis. It is not a system of government, which is what the memo’s question asked about. Six percent named it anyway.

Democratic socialism, as the Democratic Socialists of America define it in their own 2025–2026 program, means Medicare for All, universal rent control, tuition-free public higher education, a 32-hour work week, a wealth tax, and a Green New Deal with public ownership of major transportation and energy infrastructure. On governing structure, the program calls for replacing “the two-party system with a multi-party democracy through proportional representation elections,” a national popular vote for president, and an expanded House. It calls for public ownership of “the largest corporations.” It does not call for abolishing private property, nationalizing all industry, or a one-party state. On the specific charge the label carries, the DSA program asks for more parties, not one.

Social democracy describes the Western European welfare-state model, built and defended inside multiparty elections.

New Deal Democrat describes the tradition that produced Social Security and the minimum wage. Representative Ro Khanna notes that Republicans made the same charge against Franklin Roosevelt four times.

Correct use:

“Zohran Mamdani is a member of the Democratic Socialists of America.” Names the organization. Checkable.

“Darializa Avila Chevalier posted ‘seize the means of production’ in 2020 and now says she is not a communist.” Names the evidence and the response.

Incorrect use:

“The Democrats are communists.” Hides which of six positions is meant. Names no policy. Cannot be checked, which is the feature.

“Communism is on the ballot this November.” No communist party is on any 2026 ballot in the United States. The sentence transfers an alarm without making a claim.

No outrage

Three weeks have passed. A confidential memo to a congressional caucus reached the press and then the public internet. I found no statement from co/efficient, from Ryan Munce, or from the House Republican Conference about the memo, about its leak, or about whether the guidance stands.

The coverage has its own gap. CNBC ran a full piece on August 4 about why Michigan voters were hearing “communist.” The Washington Times ran one on July 28 about Republicans framing the midterms as a battle against communism. Both quote Johnson. Both quote McClain. Neither mentions the memo that told them to say it.

The instruction is public. The document is public. The strategy gets covered as though it emerged from conviction.

The numbers under the instruction

The memo’s authority rests on a survey that I do not believe anybody outside the firm has seen.

co/efficient is a real shop with a real record. Silver Bulletin grades it B+ with a moderate Republican lean. Federal Election Commission records show Republican campaigns and committees paying the firm through this cycle for polling and survey research, including $42,500 from the National Republican Congressional Committee on May 29, 2026. Nobody should read this as a front operation.

The July survey is a separate question. The memo reports 2,634 registered voters, fielded July 13 to 16, mobile text responses and landline interviews, weighted to age, gender, education, race, and region, margin of error 1.88 points. No toplines. No crosstabs. No question order. Three weeks after the document became public, none of that has been released, and the firm has not answered questions about it.

co/efficient states its own standard on its website, under a heading that reads Radical Transparency: “No black boxes. You see our methodology, our data, and our reasoning. We publish our work because we stand behind it.”

The instruction that 42 percent justifies is quoted, dated, and consistent across four outlets. The 42 percent is a number the country has been asked to take on faith.

One-minute script

When the word comes up in conversation, say this:

“That word is carrying at least six different things right now. There is communism, meaning a one-party state. There is Marxism, which is an analytical method. There is democratic socialism, which in this country means Medicare for All and rent control and proportional representation. There is European-style social democracy. There is New Deal liberalism. And there is ‘any Democrat.’ Which one do you mean? And what do you want done about it?”

Log prompt

The next time you hear “communist” from a candidate, an ad, a news anchor, or a neighbor, write down four things:

1. Who said it.

2. Which specific person or policy they applied it to.

3. Which of the six positions above that person or policy actually matches.

4. What action the speaker wants taken as a result.

If you cannot answer three, the word did its job and you learned nothing.

Read the ballot the way you read this piece. Every race, every proposition. Vote. vote411.org

Munce named the mechanism himself, on page 1, in a section heading: The Only Variable Is the Words. He proved it on a tax question, then turned it on three political labels.

There is one more thing worth knowing, and it belongs to me rather than to the memo. On September 25, 2025, NSPM-7 directed the Joint Terrorism Task Forces, the Attorney General, the Treasury, and the IRS toward domestic political violence. Its list of “common threads animating this violent conduct” includes “anti-Americanism, anti-capitalism, and anti-Christianity.” The memorandum instructs the IRS Commissioner to refer tax-exempt organizations and “the employees and officers of such organizations” to Justice for possible prosecution.

The co/efficient memo says nothing about any of that. It is campaign language, and it should be read as campaign language. But an adjective now sits in two places at once. In one, it moves a poll number. In the other, it appears on a federal list of investigative indicators. In our July 17 Binding on how an ideology becomes a Foreign Terrorist Organization, we traced how executive-branch language re-scopes institutional authority abroad. The word “anti-capitalism” is the domestic version of that same move, and a campaign that spends four months teaching a country to hear “communist” as a description of its neighbors is building the audience for it.

Ask which one they mean. Every time. Especially when the answer sounds obvious.

Sources

• co/efficient, “The Fight Over the Economic Future of America,” strategic memorandum to House Republican Conference member offices, July 20, 2026. National survey of 2,634 registered voters fielded July 13–16, 2026; ±1.88 at the 95 percent confidence level.

• Hailey Bullis, “Memo urges Republicans to tie Democrats to communism,” Washington Examiner, July 2026.

• Chris Lehmann, “The GOP’s New Red Scare,” The Nation.

• Garrett Downs, “Why Michigan voters are hearing ‘communist’ as they vote Tuesday,” CNBC, August 4, 2026.

• George Caldwell, “GOP frames midterms as battle against communism as socialists gain foothold in Democratic Party,” The Washington Times, July 28, 2026.

• Democratic Socialists of America, “Workers Deserve More!” 2025–2026 Program.

• National Security Presidential Memorandum/NSPM-7, “Countering Domestic Terrorism and Organized Political Violence,” September 25, 2025.

• MS NOW interview with Darializa Avila Chevalier.

Subscribe now

Jerry W. Washington, Ed.D., served 23 years in the Marine Corps. He holds a doctorate in Organizational Change and Leadership from USC Rossier and teaches at UC Irvine Division of Continuing Education. He writes What Time Binds, an independent newsletter on language, institutions, and democratic accountability.

Subscribe now

On May 11, 2026, Donald Trump posted this to Truth Social: “During my Historic Election in 2024, when I won every single Swing State, and decisively won both the Electoral and Popular votes by wide margins, the Republicans had an Election Integrity Army in every single State to preserve the sanctity of each legal vote.” He said the 2026 version would be “much bigger and stronger.”

The next day, the Republican National Committee formally launched the program in 17 states. The RNC did not name which 17.

Twelve weeks later, “Election Integrity Army” is a phrase running through Facebook feeds, cable panels, and state election-office phone calls. Ask three people what it means, and you will get three different answers.

That is the problem.

Share

The word

“Election Integrity Army.”

Why it is a magnet

The phrase covers at least four operations happening in parallel right now. Each has its own legal footing. Each has its own rules of behavior. All of them share the phrase.

1. RNC poll watchers. State law governs what a poll watcher can do. In most states, watchers must be credentialed by a party or campaign, cannot approach voters, and can be removed for disruption. This is the version most reporters describe when they write the phrase into a headline.

2. The RNC legal apparatus. Newsweek’s May reporting described the “operational core” of the “army” as primarily legal: RNC lawyers, election-integrity directors in 13 battleground states, and volunteers positioned to become witnesses in litigation. This version does not appear at your polling place. It appears in court after Election Day.

3. Volunteer networks organized by outside activists. Scott Presler’s Pennsylvania operation is one example. These networks recruit door-knockers, ballot-chasers, and self-described “election observers” who often are not credentialed under state law.

4. Federal law enforcement or military at polling places. The administration has refused to rule this out. Trump on May 12 told PBS he would do “anything necessary.” Acting Attorney General Todd Blanche defended deploying ICE to polls at a March conservative conference. Senator Elissa Slotkin asked whether federal troops would go to polling places; Defense Secretary Pete Hegseth dismissed the question as a “gotcha hypothetical.” On July 17, DHS Secretary Markwayne Mullin threatened state election officials with “fines, penalties, and prison time” for refusing to hand over voter records.

Four operations. One phrase. Different laws for each.

The failure mode when meanings collide

A voter walks into a polling place and sees someone in a windbreaker with “Election Integrity” printed on the back. They freeze. Which one is it? Is it a state-credentialed watcher who cannot approach them, or a federal agent who can detain them, or a volunteer with no authority at all?

The voter is left to guess. The election judge is left to guess. The poll worker at the check-in table is left to guess.

The phrase functions this way.

An election judge who reads “Election Integrity Army” as credentialed poll watchers under state observer law lets someone in the door. That person may operate under a different set of assumptions. A voter reads the same person as a threat or an authority. Not everyone is reading from the same rulebook. The phrase absorbs every version and offers no way to sort them.

When meaning drifts under coordination pressure, coordination fails. That is what happens next.

The pin

“Election Integrity Army” is a rhetorical umbrella phrase. It has no single legal meaning. It refers to at least four separate operations running in parallel: (1) RNC poll-watching programs governed by state observer law; (2) an RNC-led legal apparatus filing pre- and post-election litigation; (3) volunteer networks organized by outside activists; and (4) federal law enforcement or military deployments the administration has refused to rule out.

If someone uses the phrase, ask which one they mean before you take any action based on it.

Boundary test

Includes:

• Credentialed RNC poll watchers operating within state observer statutes.

• RNC legal teams filing lawsuits, challenging results, or documenting witnesses.

• Volunteer networks affiliated with the RNC operation.

Excludes:

• Sworn peace officers on duty for polling-place security under state or local jurisdiction.

• Nonpartisan election judges and clerks.

• Federal armed personnel at polling places. Prohibited under 18 U.S.C. § 592 (dating to 1865) and 18 U.S.C. § 594 (voter intimidation). Air Force General Gregory Guillot told the Senate Armed Services Committee on March 20 that armed uniformed personnel should not be “anywhere near” polling sites.

Correct use:

“The RNC’s Election Integrity Army poll-watching program will operate in 17 states.” Specifies version. Specifies scope.

“The Election Integrity Army’s litigation team filed suit in Georgia this week.” Specifies version. Specifies function.

Incorrect use:

“The Election Integrity Army will be at your polling place.” Hides which version. Hides scope. Hides what the person there can lawfully do.

“The Election Integrity Army will make sure only citizens vote.” Implies enforcement authority the RNC program does not have and the federal law-enforcement version has not been authorized to exercise at polling places.

Before the polling place: the registration question

If you cannot know who is standing near you at the polling place, you also cannot assume you know where to register in the first place. Which site is the official voter registration site this cycle? That question has moved. The federal government has spent the last six months trying to reshape it.

Vote.gov used to be the plain answer: the federal front door that pointed voters to their state registration portals. Its status is in question this cycle. DHS is attempting to fold voter registration and verification into the SAVE Act enforcement apparatus, and a federal court has already blocked that “updated system” on voter-privacy and wrongful-purge grounds. Until the dispute settles, the reliable official registration site is your state’s Secretary of State portal. Not vote.gov. Not any third-party lookalike. Not any site that asks for information beyond what your state’s own registration form requires. If you already registered and cannot remember whether it went through, your Secretary of State’s office will confirm your status by phone.

One-minute script

When someone raises the phrase in conversation, say this:

“That phrase covers at least four different things. There is an RNC poll-watching program — state law governs what those watchers can do. There is an RNC legal team filing lawsuits. There are volunteer networks. And there is federal law enforcement or military the administration has not formally deployed but has not ruled out. Which one are we talking about? Because the rules are different for each.”

Log prompt

When you next hear “Election Integrity Army” from a friend, a candidate, a poll worker, or a news report, write down four things:

1. Who said it.

2. What specific action they described.

3. Which of the four versions above matches that action.

4. What your state law says about that version at your polling place.

If you cannot answer version four, call your state Secretary of State’s office before Election Day.

Return to the opening

Meaning drift is an infrastructure problem. When a phrase is engineered to absorb multiple incompatible operations under one banner, every actor working near it loses shared reality. Election officials cannot coordinate. Voters cannot know their rights. Poll workers cannot enforce the rules.

In my July 17, Binding on how an ideology becomes a Foreign Terrorist Organization, I traced how executive-branch language re-scopes institutional authority abroad. The domestic version of that pathway is running now. A DHS Secretary threatened state officials with prison time using a system a federal court had already blocked. That threat did not come from nowhere. It came from a longer pattern of language that dissolves the boundary between what the federal executive branch can lawfully do and what it says it will do.

Pin the phrase. Ask which version. If you see or experience intimidation at a polling place, report it to the DOJ Voting Section at 1-800-253-3931, to Election Protection at 866-OUR-VOTE, or to your state election board.

Before you cast your ballot, know what “Election Integrity Army” means the person standing near you is actually allowed to do.

Subscribe now

Sources

• Trump Truth Social post, May 11, 2026

• RNC operation launch, May 12, 2026 (Democracy Docket, Newsweek, The New Republic)

• Trump “anything necessary” quote, PBS Newshour interview with Liz Landers, May 12, 2026

• Heather Honey (DHS Deputy Assistant Secretary for Election Integrity) call with state election officials, February 25, 2026 (NBC News, CNN)

• Air Force General Gregory Guillot testimony, Senate Armed Services Committee, March 20, 2026 (Democracy Docket)

• Acting AG Todd Blanche remarks on ICE at polls, March 2026 conservative conference

• Defense Secretary Pete Hegseth “gotcha hypothetical” dismissal, Senate hearing, April 2026

• DHS Secretary Markwayne Mullin news conference, July 17, 2026 (New York Times, PBS NewsHour, Spectrum News, The Hill, NBC News)

• Brennan Center for Justice, “Sending ICE to Polling Places Is Illegal,” March 31, 2026

• Padilla and Slotkin legislation, July 2026

• DNC lawsuit against DOJ, Pentagon, DHS, July 2026 (NOTUS)

• 18 U.S.C. § 592, § 594

Subscribe now

Sometime in the mid-1970s, a sergeant in the 82nd Airborne Division got word that a general was coming to see him. Not his unit. Him.

The sergeant was Bruce Watts. He had a secret he guarded carefully: his father was a general officer, and Bruce wanted to be judged on his own merit, so he told no one. The visiting general had eaten dinner at his parents’ table a few days earlier and promised to look in on their son. So Maj. Gen. Charles Calvin Rogers appeared in front of a terrified young NCO, shook his hand, chatted with his troopers until his handlers pulled him away to see the division commander, and left a chain of command spending the rest of the day asking who Bruce’s father was.

Years later, Bruce read the Medal of Honor citation. Fire Support Base Rita, Vietnam, November 1968. A lieutenant colonel wounded three times who kept directing fire, kept leading, and kept his firebase from being overrun. Bruce wrote something about that citation over the weekend that has stayed with me since. Heroism like that, he said, “must be published and displayed in conspicuous places” to inspire the young soldiers coming up.

Conspicuous. Hold that word. The citation’s own phrase is “conspicuous gallantry.” Conspicuous is the standard the record was written to meet.

Now the word this Field Guide pins.

Sixteen months ago, the Defense Department’s tribute page for General Rogers vanished from the Pentagon’s website. The replacement URL contained three letters: dei. The removal drew cameras, then backlash. Three days later the page was back, and the story everyone remembers ends there: restored. Case closed.

“Restored” is the word that ended the attention. Here is what the datestamps say happened while the word did its work.

March 15, 2025: an independent archive catches the page mid-removal, returning a 404 error. March 17: the page comes back. March 18: a capture of the returned page shows an edit already inside it. “Gender and race equality” now reads “sex and race equality.” Processes don’t copyedit. The page everyone celebrated as restored had been changed before the celebration finished.

March 21: the Pentagon issues a policy statement. Removed content is under review “to determine if it was mistakenly removed or, in other cases, if it can be edited to make it eligible for reposting.”

Read that twice. Wayne Coy, whose guest essay ran here Monday, read it once and named it: “The standard was not accuracy. It was eligibility.” Edit the record until it clears the filter, post it back, call it restored.

By the April 5 capture, the editing was done. Same headline. Same byline. Same 2021 dateline. No correction note. Five sentences gone:

“As a Black man, he worked for gender and race equality while in the service.”

“Rogers, who attended the all-Black Dubois High School during the segregation era, excelled as a student.”

“His first few years as a soldier were spent serving in artillery commands while the service was being desegregated.”

“To this day, he is the highest-ranking Black Medal of Honor recipient.”

“Rogers has been remembered not just as being a Medal of Honor recipient, but for championing gender and race equality during his long tenure in the Army.”

The Defense Department published every one of those sentences in 2021, under its own byline. They stood for three and a half years. Then they left without a note, from a page wearing a dateline that swears nothing happened.

So, what do you mean, “restored”?

The dictionary means returned to its former state. The page returned to visibility. The record never returned at all. What came back is a document that looks like the original, carries the original’s date, and says less than the original said. The word covered the difference. That is the whole function the word performed: it told the country the story was over while the story kept going.

I posted the receipts on Facebook on Saturday night. As of this morning, the thread holds 264 comments, and what happened inside it is the reason this entry exists.

The Pentagon removed five sentences. The thread produced five witnesses.

Bruce told the handshake story. A soldier wrote that he recovered beside Rogers in a hospital in Japan, both of them wounded, playing hours of cribbage. An officer remembered Rogers speaking at his Ranger School graduation in 1977, before he knew what the man had done at Rita. Another reader served under him as a young lieutenant in the mid-70s and walked the thread through the direct-fire decision at the firebase, gun by gun. An artilleryman wrote that Rogers is the reason he chose the branch, and pointed to the howitzer that stood with him at Rita, a rocket hole in its side, displayed outside Snow Hall at Fort Sill.

Five men, five different decades of contact, surfacing on one post inside a day. One reader named the tradition at work: the griot, the keeper who carries a community’s stories when institutions won’t. Another described her practice at Arlington, laying an extra bouquet on the grave of someone being erased. Her word for it: un-erase.

This is what I mean when I write about time binding. A civilization passes its knowledge forward through substrate: pages, plaques, databases, citations. Whoever edits the substrate edits the inheritance. The Rogers page proves the substrate can be rewritten in the dark. The thread proves the inheritance has a backup system, and the backup system is people.

One more thing happened in that comment section. Several readers asked, some sincerely and some with folded arms, why the man’s race should matter at all. Why not honor the officer and skip the Black officer?

The five deleted sentences are the answer. A segregation-era high school. Artillery commands during desegregation. The highest-ranking Black recipient in the medal’s history. Each sentence records a fact about what this country asked one soldier to overcome while he served it. The question and the edit want the same thing, and the edit already acted on it. Whoever cut those sentences answered “why does race matter” with a delete key and left no name on the work.

And here is where the story stops being finished. In the thread, one reader asked whether the plaque is still on that howitzer at Fort Sill.

Nobody knew.

That is the condition the quiet rewrite leaves us in. Every marker now has to be checked. The page, the plaque, the archive, the database. Friday’s essay returns to this pattern at a memorial in Philadelphia, where a court ordered a restoration and got one, briefly. The pattern runs in three steps: removed, restored under pressure, quietly replaced. Once you can read the steps, you will see them running elsewhere.

Until then, one habit and one assignment.

The habit: when an institution tells you something was restored, ask to see what came back next to what was there before. The datestamped captures for this story are linked below. Repair needs blueprints, and blueprints are what archives are for.

The assignment: an edit history is a public record. So is your voter registration, and the people who hold the pen over both get chosen at the ballot. Check your registration the way you would check that plaque. Read every line. vote411.org.

The record of Charles Calvin Rogers survived because people carried it. Carry it.

Share

Receipts, datestamped:

My March 18, 2025 piece, written while the page was down:

https://medium.com/@jerrywash/trumps-weakness-exposed-4ea0f66687ad

My Facebook Post: https://www.facebook.com/jerrywashington.edd/posts/pfbid0g8Rn3UpVYabBN8GsqLM7rzgrsaT2kigFx4jTAiLhcjgcKL623aYWnKJ3df1dkhrRl

What Time Binds studies how meaning drifts and how we repair it. One field guide like this every week, free.

Subscribe now

Subscribe now

The word

DEI. Three letters that used to be a shorthand for a portfolio of workplace practices and civil-rights obligations. Now the letters are a battleground.

Why it’s a magnet

Many definitions travel under those three letters. Each carries the weight of the others. Each speaker acts as if their definition is the only one on the table.

Below are examples I dealt with this week.

Definition 1: DEI as segregation. Racial exclusion enforced by law or organizational fiat. This is Thomas Guastavino’s opening move on a Facebook thread I posted this weekend: “Does the author know DEI is a form of segregation?” Definition 1 requires collapsing Jim Crow (state-mandated racial exclusion) with Title VII (voluntary access-widening under statute). The distance between the two is the entire civil-rights record of the last sixty years.

Definition 2: DEI as unconstitutional preference. Race-conscious action that violates the 14th Amendment. This is Tom Kelly’s opening move: “Dei violates the 14th amendment.” Definition 2 requires treating cases about specific doctrinal areas (university admissions in SFFA v. Harvard, 2023; government set-asides in Croson 1989 and Adarand 1995) as if they had ruled on DEI as a whole practice. None of them did.

Definition 3: DEI as weaponized advantage. A mechanism by which someone who did not earn the job forces the employer to hire them. This is Ryan McCoy’s move: “Sadly some people that got an interview but didn’t make the cut used DEI to force them to get hired...” Definition 3 requires an unnamed anecdote to stand in for a documented case.

Sitting under all three, the actual referent: DEI as a portfolio of statutory obligations and institutional practices. Title VII (1964). ADA (1990). VEVRAA (1974). Section 503 of the Rehabilitation Act (1973). PWFA (2023). EEOC barrier analysis under 29 CFR 1608 and 41 CFR 60. On top of the statutes: outreach, mentorship, employee resource groups, bias training, standardized rubrics.

The failure mode

Many definitions in one word means every debate about DEI is at least a few debates layered on top of each other, with participants speaking past each other while agreeing the word matters. The Facebook thread I ran this weekend and the House Oversight hearing on Tuesday morning were the same failure mode at different scales. Each speaker used DEI to attack a version nobody defended and defend a version nobody attacked.

The failure resolves in one direction. Whoever is asked to define what they mean, and refuses, gets to keep the ambiguity as an argument.

Share

The pin

DEI, as it exists in current law and institutional practice, is a portfolio of statutory obligations covering race, sex, disability, veteran status, religion, and pregnancy, plus voluntary practices that expand recruitment, reduce evaluator bias, and support retention. Race-decisive decisions and hard quotas are the illegal edge case, not the portfolio.

That definition is testable. It survives contact with the statute. It survives contact with EEOC guidance. It survives contact with the case record.

Boundary test

Includes:

• Barrier analysis under EEOC MD-715

• Disparate-impact enforcement under Griggs v. Duke Power (1971)

• ADA accommodations for qualified applicants with disabilities

• VEVRAA and Section 503 outreach to veterans and people with disabilities

• Mentorship, sponsorship, and employee resource groups

• Standardized evaluation rubrics

• Pregnancy accommodations under PWFA

Excludes:

• Hard numerical quotas (illegal in employment since United Steelworkers v. Weber, 1979; illegal in admissions since SFFA v. Harvard, 2023)

• Race, sex, or another protected class as a decisive factor in a specific hiring decision (illegal under Title VII disparate-treatment doctrine since 1964)

• Race-conscious plus-factors in university admissions (illegal since SFFA v. Harvard, 2023)

Example. A federal contractor conducting a barrier analysis under 41 CFR 60 to identify neutral policies that produce disparate outcomes. That is DEI.

Non-example. An employer refusing to hire white applicants. That is a Title VII violation. Title VII has said so for sixty-two years.

One-minute script

When someone tells you DEI is discrimination, or DEI violates the 14th Amendment, or DEI is a form of segregation, ask three questions. Take about a minute total.

1. “Which DEI practice specifically? Barrier analysis? ADA accommodations? Veteran outreach under VEVRAA? Employee resource groups?”

2. “Which employer ran the practice, and which mechanism in it produced the harm you’re describing?”

3. “Which statute, regulation, or court ruling defines that practice as unlawful?”

If the answer arrives, engage the answer on its merits. If the answer does not arrive, the silence is the argument. File the silence next to the claim.

Log prompt

Before you use DEI as a target or as a shield, log two things.

1. Which of the three definitions am I using: segregation, unconstitutional preference, or weaponized advantage?

2. Which specific DEI practice am I pointing at?

If you cannot answer the second question, you are arguing about a category. Categories don’t hire. Categories don’t accommodate. Categories don’t approve or deny a home loan. Programs do those things. Name the program.

This is a Field Guide entry. Companion to today’s special essay, Name the Mechanism, which traces the same pattern through three Facebook commenters and the House Oversight hearing on Tuesday, July 14. Both pieces are the record built on one Facebook photograph and one CSPAN clip.

Related WDYM entries: What Do You Mean, “Allegiance”? — where “DEI Justice” showed up as a smear-magnet. When “Award” Stops Meaning Award — same repair pattern applied to a different magnet word.

Subscribe now

Sources

• Statutes: Title VII of the Civil Rights Act of 1964; Rehabilitation Act of 1973 §503; VEVRAA of 1974; ADA of 1990; Civil Rights Act of 1991; PWFA of 2023.

• Regulations: 29 CFR 1608 (EEOC affirmative-action guidelines), 41 CFR 60 (OFCCP contractor obligations), EEOC MD-715 (federal employer barrier analysis).

• Cases: Griggs v. Duke Power (1971), Regents of the University of California v. Bakke (1978), United Steelworkers v. Weber (1979), City of Richmond v. J.A. Croson (1989), Adarand Constructors v. Peña (1995), Grutter v. Bollinger (2003), Students for Fair Admissions v. Harvard (2023).

• Facebook thread source material: WTB Essays/Hayakawa Series/_Source Materials/2026-07-12_Dhillon-Thread-Case-Study.md.

• House Committee on Oversight and Government Reform, Task Force on Defending Constitutional Rights and Exposing Institutional Abuses, Combating DEI in American Institutions, Tuesday, July 14, 2026, 10 a.m. ET. Chair: Brandon Gill. Witnesses: Mike Gonzalez (Heritage Foundation), Inez Feltscher Stepman (Independent Women), Michael Shires (America First Policy Institute).

Subscribe now

After hundreds of masked men marched through Washington on July 4, Laura Ingraham posted a video of the column and wrote: “I call fake. Looks more like Antifa in costume.” Then she added a rule: no one should be allowed to cover their faces.

There was snow on the ground in her clip. July 4 in Washington was a heat evacuation day. Old footage, new job.

The next morning, I posted the fact pattern. Hundreds of masked men. Confederate and inverted US flags. “Reclaim America” chants, documented by Washington Post, NBC4, WTOP, and Reuters photographers. Ingraham agreed all of it happened. She reassigned the actors.

The post drew over 400 comments in two days. I treated the thread the way a researcher treats returned surveys. I captured 416 entries and coded them into fifteen themes, one primary theme per comment, following Braun and Clarke’s standard method for thematic analysis. One coder (me), a purposive capture, counts offered as illustration rather than findings. I want to be honest about what this essay is.

Here is what the coding showed. Every section of a field guide entry, the ones I usually write on Wednesdays, already existed in that thread. The commenters wrote the magnet analysis, ran the failure mode live, supplied the pin, and drafted the scripts. Four hundred strangers performed the repair protocol without ever having read one.

So this essay runs inverted. The crowd is the author. I’m the editor with a codebook.

Share

The word

Everyone in the thread fought over “antifa.” Thirty-nine comments pinned it: no headquarters, no membership rolls, a word built from “anti-fascist,” a lineage that runs back to men who landed at Omaha Beach. Janine Kravetz gave the cleanest version. “What does Antifa look like? It looks like my father, serving in WWII in Italy.”

The load-bearing word sat one position to the left.

“Costume.”

Watch what it does. “Costume” concedes every visible fact and reassigns every invisible one. The march happened. The men were there. The flags flew. Ingraham granted all of it, because “costume” let her keep the footage and swap out who was inside the clothes. The word converts eyewitness record into wardrobe.

A ladder for this was created seventy years ago to help us understand: a report, then an inference, then judgment. A report can be verified. “Masked men marched in Washington on July 4” is a report. “Antifa in costume” is a judgment sprinting past the report. That’s the whole game, and “costume” is the word that is in play.

Why it’s a magnet

A magnet word pulls a room into different meanings. “Costume” pulls harder than most, because once the visible is declared a disguise, anyone can be underneath.

The thread proved this by rotation. Seven different labels passed through the same slot in two days: Antifa. Democrats in disguise. Feds. SPLC hires. Paid actors. A psyop. One commenter went with Hamas. The labels never agreed with each other. The slot never moved.

One commenter in the thread saw the mechanism from inside it. Michael Dixon proposed the mirror image, that violent “antifa” could just as easily be right-wingers cosplaying to build their own boogeyman, then said the quiet part plainly: with masked guys, you could claim they are anyone. That is the magnet. The mask plus the word “costume” produces a claim no evidence can touch, because every face you can’t see is available for reassignment.

The failure mode, run live

In 2010, political scientists Brendan Nyhan and Jason Reifler published a study called “When Corrections Fail.” Their finding: on charged questions, corrections often leave the misperception standing. Sometimes they strengthen it.

My comment section replicated the study for the second time in a month. In June, the thread under my post about General Chappie James’s removed portrait produced the same pattern, coded and archived. This time the specimen was cleaner.

One commenter reasoned his way to the relabel: why would Trump supporters march against a sitting Trump government? Sounds like antifa. Another commenter answered with content, the group’s own manifesto and its “reclaim it for white people” aim, flags included. The correction landed. The first commenter updated: seems like a ploy to sow division, I bet they’re all paid.

Read that sequence again. The evidence arrived. The conclusion survived by changing costumes. Antifa exited, paid actors entered, and the underlying claim, that the men on camera cannot be who they say they are, never blinked. Corrections kill labels. The slot regenerates.

Then the failure mode did something I did not have in June’s data. It jumped hosts.

The photograph

On the same train those men rode, Reuters photographer Cheney Orr made a picture: a young Black woman seated alone, her face uncovered, surrounded by masked members of Patriot Front. By Sunday it was one of the most shared images in the country, with commentators predicting it will sit in history textbooks. Her family, reached by the Washington Post, asked the public to remember she is “more than a symbol.” A second photograph, by Finn Gomez for Getty, shows a Black man in the same position. A third rider, Roswell Encina, posted his own account of sitting on that train for twenty-five minutes.

Within a day, posts claimed the Reuters photograph was staged or AI-generated. Snopes ran it down: authentic, no manipulation, no AI, matched to the original on the Reuters wire.

Look at what just happened. The costume claim did not work, so it moved to the pixels. First the marchers were somebody else in disguise. Then the photograph was software in disguise. Same slot, new object. Once “costume” is admitted as an explanation, nothing visible is safe from it, including the camera itself.

Reuters captioned the woman with two words the internet skipped past: “A commuter.” A person going somewhere, on a public train she paid to ride. The caption is a report. Everything piled on top of her since, the Rosa Parks comparisons, the Pulitzer talk, the fakery claims, is inference and judgment doing what they do. Her family’s four words are a repair: more than a symbol. They are asking the country to climb back down the ladder.

Pin it

In this situation, “costume” means a claim that the visible record misrepresents its own participants. The claim is legitimate exactly once it carries its burden: name one person under the mask who belongs to another group.

Includes: an identified infiltrator, a documented plant, a named actor with a paper trail. Excludes: vibes, motive puzzles, “looks like,” and every explanation that requires the mask to stay on.

The boundary test walked past everyone on camera. Thomas Rousseau, who founded Patriot Front, marched unmasked at the front of that column, directing chants, on video, photographed by the same wire services Ingraham’s network subscribes to. Several commenters caught it independently. One put the behavioral test perfectly: if the marchers were really someone else, the unmasked founder would have started a brawl.

And identity was always recoverable. In June 2022, police in Coeur d’Alene, Idaho pulled 31 Patriot Front members out of a single U-Haul near a Pride event. The booking photos are public. Masks come off in custody. One identified face settles what a thousand masks obscure, and the faces have been identified before, by name, under arrest.

That’s the pin. Hold it and the word “costume” has to show its work or leave.

What the crowd already knew

The finding that matters most for this series: nobody taught these commenters the protocol. They ran it anyway.

Fifty-nine of them, the largest block of substantive argument in the thread, independently built what I’d call a ledger. Hold the speaker’s rule against the speaker’s own applications. “No one should be allowed to cover their faces” met ICE within minutes, dozens of times, from dozens of unconnected people. Seventeen more ran consistency tests: if those men are antifa, a designated terrorist organization under this administration, where are the arrests? DC police reported none. No complaints, no calls for assistance. The claim collided with its own consequences and the commenters kept score.

Others dated the footage by the snow and the parkas. One dated a substitution photo posted in my comments to a specific 2020 rally in Louisville within hours. One read the Roman fasces on the marchers’ own flags back to the root of the word “fascism.” Jeff Troxel wrote a complete repair script in eleven words without knowing the genre existed: “’Looks’ like Antifa? What does Antifa look like? Is there a description?”

The repair protocol is folk practice. People already make these moves under pressure. The field guide names them so they can be made on purpose, earlier, and out loud.

One more finding, and the entry is dishonest without it. The single largest theme in the thread was the pile-on: 68 comments of insult, much of it aimed at Ingraham, a heavy share of it plainly misogynist. The same thread that dated the footage and named the fasces also produced sexual degradation as a form of argument. The immune system generates inflammation. Correction and contempt traveled in the same replies, sometimes in the same sentence. Naming the mechanism means naming it on every side of the thread, including mine.

The protocol

Ten prompts, drawn from what the thread did right.

1. What do you mean by “costume”? Who, specifically, is inside it?

2. Name one person under the mask. The claim owns that burden, permanently.

3. Does your explanation survive the unmasked man at the front of the column?

4. Which label are you on? If it changed since your last comment, the label was never the claim.

5. Write down what was conceded. She agreed the march happened. Keep that on the record.

6. Hold the rule against its author. Who else covers their faces, and did you object then?

7. If the label were true, what would follow? Arrests? Denunciations? Are they happening?

8. Date the footage. Weather is a timestamp.

9. Read the symbols the group chose for itself. Flags are signed confessions.

10. Separate the report from the judgment before you reply. Answer the report first.

The one-minute script, for the next time “costume” shows up in your feed or your meeting: “You’ve agreed the event happened. So the only claim left is who was inside the clothes. Name one. Until then, the record stands as filmed.”

Log it

Write the definition down where future-us will find it. This time the thread did that too. Metro timestamps, photo metadata, a snowbank in July, booking photos from Idaho, a Reuters caption, and 416 coded comments now sitting in a spreadsheet with the date on it.

The relabel needs the record to be soft. This record set like concrete while the labels were still rotating.

Every masked man in that car is now the subject of a costume claim, his own group’s or Ingraham’s. Between those two claims sits a woman on a train whose face required no argument at all. Her family already gave us the only caption she needs.

More than a symbol. A commuter. She paid her fare and kept her face.

Subscribe now

Previously, on What Time Binds: At the Level of Words, At the Level of Facts opened the Hayakawa three-part series with the 2026 inversion of his 1953 diagnosis. What Do You Mean, “Rally Around the Flag”? ran the same diagnostic on a different magnet word. Feynman’s Real Question Was “What Do You Mean?” is where this series’ method got its name.

What word is doing costume work in your feed this week? Name the label and the slot in the comments.

Subscribe to What Time Binds for Friday’s Binding, and Monday’s Hayakawa Part 3: the testimony log.

Methods note: 416 comments captured July 5–7, 2026 and coded by a single coder using thematic analysis (Braun & Clarke, 2006). Purposive capture, counts as illustration. Sources: Reuters (Cheney Orr photograph and caption, July 4, 2026); Snopes fact-check on the photograph’s authenticity; Washington Post, July 7, 2026 (family statement); The Advocate (Roswell Encina account); ADL and GWU Program on Extremism backgrounders on Patriot Front; Nyhan & Reifler, “When Corrections Fail,” Political Behavior 32(2), 2010; Kootenai County, Idaho arrest records, June 11, 2022.

Subscribe now

Kuwait, 2003. We staged for the push into Iraq and drew a mission to secure an airfield. The intel imagery showed debris across the field. Cars, set in places too deliberate to be random. We read them the only responsible way. Vehicles staged like that get wired. We planned the clearance on that assumption and treated every car as a live charge.

We reached the field. The cars sat where the imagery showed them. No mines, explosives, or anything wired.

Here is what has been in my mind ever since. The debris was in place, and there were no explosives. We prepared for what could have happened. I will never know whether we prevented a mass-casualty event or whether there was never a bomb to prevent. The disaster that does not arrive cannot tell you whether it was coming. If there was something there, we never found it.

That is the trouble with the word “prevented.” It carries two claims, and they are not the same.

One claim says I lowered the odds. You can check that. The other claim says without me, the bad thing happens. You really can not check that, because the event that never occurred leaves nothing to measure. Most arguments about prevention are two people holding different halves of the same word. If we want to discuss the same thing, we need to ask each other more questions.

Share

Put the index back

Monday’s essay drew Hayakawa’s split between the level of words and the level of facts. “Prevented” is where that split shows up inside a single word. Hayakawa taught a second tool for exactly this: the indexing rule. Chair-1 is not chair-2. Two things wearing the same word are never identical. The word hides the difference until you add the number.

Prevention-1 is a single averted event. The airfield. One day, one field, one charge that may or may not have been real.

Prevention-2 is a lowered rate of events across many cases. Insurance lives there. So do vaccines, seatbelts, and rehearsed procedure. You cannot prove the policy saved you this month. You can prove, across thousands of policyholders, that the pooled prevention is real and priced to the dollar.

The word “prevented” drops the index. Dropping it is where problems arrive.

Prevention-2 from the same war

We ran convoys in Iraq. Before we rolled, we drilled what would break. A vehicle breakdown. A truck separated from the column. Hostile fire. A tire change under load. An overheated engine. An improvised explosion. A vehicle lost from the convoy entirely. We set rally points along every route and ran the procedure for each failure until the procedure was reflex.

On one movement, a vehicle got separated. The crew did what we had practiced. They returned to the last rally point and held. We found them. Their truck had broken down, the exact failure we had rehearsed.

The event happened. It stayed small. It stayed small because we had run the repair hundreds of times before the day we needed it.

You cannot replay one convoy to test one drill. The countable claim runs across reps and units: rehearsed repair contains escalation. Run the repair enough times and the escalation rate drops, and the drop is something you can measure. The repetition was the infrastructure. The practice lowered the threshold for catching the failure before it spread. This is measurable across time and events.

Why the word gets abused

When the only thing you can point at is nothing-happened, you cannot tell the person who prevented the problem from the person who says they did. Both point at the same empty field.

A writer I read closely, , mapped this extremely clearly. Arimitsu works front-of-house in hospitality and writes from the floor. In a recent essay, they laid out why prevented work resists scoring. You cannot separate “you fixed it” from “it was never going to fail.” Put a reward on invisible work, and the first thing you summon is people who fake invisible work. That is a correct observation. The single averted event resists proof, and the faker hides in that gap.

The essay lives entirely in prevention-1. Prevention-2 is different, and it is where the measurement waits. Index up, and the seeing becomes ordinary. Actuaries do it every day. So do the people who price your car insurance and the units that drill rally points before they need them. His conclusion, that any system trying to credit prevention ends up believing what it cannot see, holds at the level of the single case. It dissolves one index higher.

Pin it in this room, today

Prevention-1 — a single averted event. You are owed a probability, not a proof. Unprovable case by case.

Prevention-2 — a lowered rate of events across many cases. Measurable, falsifiable, priced.

Includes: a drill that turns a breakdown into a delay; a policy that lowers a population’s loss rate; a rehearsed procedure that contains a real failure.

Excludes: “I stopped the disaster” when no disaster arrived, and no rate was measured. That is a sequence cosplaying as a cause.

Example: the convoy. The breakdown came. The repair held. The escalation rate dropped, and you can count it across events.

Non-example: the airfield, claimed as a save. We acted. Nothing happened. Honest reporting stops at “we prepared.” The causal claim stays unearned. We did not prevent anything, and nothing happened.

The one-minute script

When you hear “I prevented that,” or catch yourself saying it, ask which prevention is on the table.

For the single case: “What would have happened, and how do you know?” If the answer is a story about a future that never arrived, the speaker is owed a probability, not a medal.

For the rate: “Across how many times, and what changed?” If the answer is a count and a measured drop, that is the real thing, and it deserves the credit.

Then write it down where the next person will see it. Prevention-2 leaves a record only if someone makes one. The convoy drill held because someone logged the rally points before the truck broke down. The repair was written before it was needed.

Back to the field

I think about the airfield more than the convoy. The convoy I can defend with a count. The airfield I cannot. We geared up for a threat that never happened, and I will go to my grave not knowing if we prevented anything. That is OK.

That uncertainty is honest. It is also the exact spot where the word “prevented” gets worked, because the person who wants credit for the empty field sounds identical to the person who earned it. The only way to tell them apart is to ask which prevention they mean. The answer tells you whether they are owed proof or owed a probability.

Ask the question. The word will not ask it for you.

If this landed, the work goes deeper.

Every essay here draws from an active advisory practice. I work with learning and development directors, workforce administrators, and education leaders on the same problem from the other side: not the analysis, but the repair. Diagnostics, workshops, and standing advisory engagements through jerrywwashington.com. A 30-minute call costs nothing.

Jerry W. Washington, Ed.D., is a Marine combat engineer veteran, independent researcher, and instructor at UCI Division of Continuing Education. He writes What Time Binds on Substack.

Sources

Hayakawa, S.I. (1949). Language in Thought and Action. The indexing rule and the report–inference–evaluation distinction.

Arimitsu (2026). The Office of Being Disbelieved and earlier writing on the invisibility of preventive work. Published on note (Japan).

Washington, J.W. (June 29, 2026). At the Level of Words, At the Level of Facts. Part 1 of three. What Time Binds.

There is a kind of officer I learned to recognize in my twenty-three years as a Marine. Captains, usually. Sometimes majors. Sharp. Articulate. Physically squared away. They could brief a slide deck or run a range with equal confidence. They knew their people down to the radio call signs. They believed the institution worked the way they had seen it work from where they were standing, which was somewhere at the unit scale.

They were sharp at the level they had reached. They had not yet been through the levels that change you. High level staff, where you learn that nothing crosses your desk clean. Brigade, where you learn that a good plan written at midnight gets executed by tired humans at dawn. Joint, where you learn that the Army and Air Force speak different languages, run different clocks, and answer to different chains. Combatant command, where you learn that the country you are about to fight in has its own clock, its own grievances, and its own veto power over your plan.

What changes you across those levels is contact. You learn that the JAG who keeps asking the difficult question is the brake. You learn that the senior NCO who tells you the plan will not work is your forward observer. You learn that the colonel above you is reading three boards you cannot see.

The officers who have not made those passes through the levels carry a frame. The frame is real. It served them well at the rank they reached. The problem shows up when someone hands them an authority their frame cannot hold.

I am watching one of those officers run the “Department of War.”

Subscribe now

The Word, and What It Selects For

When that kind of officer gets four-star authority and goes to work on a word, something specific happens to the word. The word starts doing two jobs. It still carries the meaning it had in public: the dictionary meaning, the speech meaning, the meaning everyone understands. It also picks up a second meaning, written in the personnel decisions, the promotions blocked, the names removed from lists. That second meaning is the operational meaning. It is what the word actually selects for.

You can watch this happen with "warrior ethos."

Pete Hegseth made warrior ethos the signature phrase of his Pentagon. He used it as his stated number-one priority on day one. He built his 2024 book around it. He flew roughly 800 generals and admirals to Quantico on a week's notice in September 2025 to lecture them on it. He renamed the Defense Department the Department of War to honor it. The word has done more work in his tenure than any other.

On June 23, 2026, his Pentagon forced out General Christopher Donahue.

Donahue: West Point. Career Ranger. Former Delta Force commander. Former director of operations at Joint Special Operations Command. Commander of the 82nd Airborne Division. Combat tours in Iraq, Afghanistan, Syria. The last American soldier off the ground in Kabul, August 30, 2021. Most recently the four-star in command of U.S. Army Europe and Africa. Retired General Tony Thomas, who led U.S. Special Operations Command, called him "the finest officer I ever served with."

By every published standard for warrior ethos, Donahue clears the bar. He clears it cleanly. He clears it the way the dictionary meaning of the term would clear it.

He was forced out eighteen months into a four-star command. Effective July 2.

So the word is doing two jobs.

Stated standard: warfighter with combat record.

Operational standard: warfighter who agrees with the Secretary.

The first definition keeps Donahue. The second removes him. The Pentagon that invented the first definition is the one operating under the second.

The Frame Behind the Decisions

Why does the word drift like that? Why does the operational definition come unstuck from the stated one?

This is where the diagnostic frame from earlier earns its keep.

Hegseth reached the rank of major in the Army National Guard. Major is a fine rank. It is also the rank where you still see the institution from where you are standing: your unit, your tour, your immediate command. He served at Guantanamo, in Iraq, and in Afghanistan, commissioned as an infantry officer in 2003. He did not run a brigade. He did not sit on a joint staff. He did not command a division, a corps, or a combatant command. He left the Guard, went to Fox News in 2014, and stayed there for a decade. When he arrived at the Pentagon in January 2025, he arrived with the frame he had carried out of the Guard, into cable television, and back through the front door of the building he had spent ten years criticizing from outside.

That frame is observable in the Quantico speech.

On September 30, 2025, Hegseth ordered every general and flag officer at one-star and above to fly to Quantico with a week's notice. About 800 of them stood up from their commands around the world and made the trip. The Pentagon spent several hundred thousand dollars in flights. When the auditorium was full, the Secretary of War told them about beards. About body fat. About "no more beardos." About "fat generals." About fitness tests scored to a male standard. He told them that if his words made their hearts sink, they should resign.

That is the speech of a junior officer pitched at the four-star register. It is the speech of someone whose primary frame of military leadership is grooming standards, PT scores, and the moral certainty of a person who has not yet been forced to sit across the table from their own bad assumption. A senior leader's concern at that level is strategy, sustainment, alliance management, joint integration, succession planning, civil-military relations. The Secretary stayed on appearance.

The generals sat stone-faced. The Chairman of the Joint Chiefs reportedly told them, before the cameras turned on, to treat the speeches the way they would treat a State of the Union. Show no emotion. Keep the institution apolitical.

That is the mechanism. A leader operating below the developmental level of his role, with no internal correction available, surrounded by deputies he chose for agreement rather than counterweight.

Call the maturity gap what it is: a developmental fact. Frames mature through contact across levels, and he had not had that contact.

The responsibility scales up. The Senate that confirmed Hegseth 51 to 50, the White House that nominated him, the political coalition that pushed for him — those are the institutions with the maturity problem. They placed a junior-officer frame in a four-star decision seat. They refused to install the structural guardrails that would have caught what the frame cannot see. That is adolescent appointment-making applied to grown-up consequences. I have a name for the pattern in my own work. I call it the Adolescent Polycrisis. Tool-power outpacing the wisdom to govern it.

The Year, Documented

Five SecDefs called this last year.

On February 27, 2025, William Perry, Leon Panetta, Chuck Hagel, James Mattis, and Lloyd Austin sent a joint letter to Congress. Mattis was Trump's first Defense Secretary. They condemned the firings as reckless. They warned that the dismissals raised troubling questions about politicizing the military. They named three forward warnings:

1. That talented Americans might stop choosing military service if they would be held to a political standard.

2. That those currently serving might grow cautious of speaking truth to power.

3. That public trust in the military could begin to wither.

They asked the House and Senate to hold immediate hearings. They asked the Senate to refuse confirmation of new DoD nominations until the firings were explained. They closed with one sentence: "We're not asking members of Congress to do us a favor; we're asking them to do their jobs."

Hegseth himself supplied the clearest evidence of what he was doing. He explained on the record why he fired the senior Judge Advocates General across the services. He said they were potential "roadblocks to orders that are given by a commander in chief." That sentence is the field guide entry. The institution's primary internal check on the lawfulness of orders is, in the Secretary's stated view, a roadblock to be removed.

Then came the record.

February 2025: CQ Brown, the Joint Chiefs Chair. Lisa Franchetti, the first woman Chief of Naval Operations. Jim Slife, the Air Force vice chief. Linda Fagan, the first woman Commandant of the Coast Guard. Shoshana Chatfield at NATO. The senior JAGs across services.

April 2025: Jennifer Short, Hegseth's own military assistant.

August 2025: David Allvin, Air Force chief of staff. Jeffrey Kruse at the Defense Intelligence Agency, fired after his agency produced an Iran strike damage assessment that contradicted the president's claims. Rear Admiral Milton Sands at Naval Special Warfare.

Fall 2025: Alvin Holsey resigned at SouthCom, reportedly over concerns about strikes on civilian-targeted vessels in international waters.

April 2026: Randy George, the Army chief of staff, confirmed 96 to 1 by the Senate in 2023.

May 2026: Hegseth personally removed nine officers from the Navy rear admiral promotion list. Three women. Two Black men. He blocked nine Air Force promotions the same month.

June 23, 2026: Donahue.

About two dozen senior officers gone in seventeen months. Sixty percent of those removed are female or Black, in a population where women and minorities make up roughly twenty percent of generals and admirals. Senator Jack Reed entered the figure into the record. This does not include the block promotions.

The Reagan Institute publishes an annual survey of public confidence in the military. In 2018, the number was 70 percent. In December 2025, the number was around 50 percent. Confidence among Democrats: 33 percent. Among Republicans: 67 percent. The partisan gap widened sharply on Hegseth's watch. The SecDefs' third warning, about trust beginning to wither, has become data.

What Future-Us Inherits

The captains and majors are watching.

They are watching what gets you removed. They learned in August 2025 that you do not produce intelligence that contradicts the political layer. They learned in September that you do not raise concerns about civilian-targeted strikes. They learned in April 2026 that you do not protect officers from a political block of promotion. They learned in June 2026 that you do not be a four-star with a textbook combat record and a public spine. They learned in February 2025, on the record, in Hegseth's own words, that you do not be a JAG who treats the law as a constraint on a commander in chief.

Those lessons are being absorbed now, by officers who will not reach four-star authority for twenty years. The selection is happening at the bottom of the pipeline. The four-star corps of 2046 is being chosen in the captains who decide right now what to say in the morning meeting.

What future-us inherits is a senior officer corps composed of the people who passed those tests as captains. The ones who learned to read the room. The ones who learned to keep the difficult question to themselves. The ones who learned that selection rewards agreement over judgment.

That is the cost the SecDefs were trying to prevent. They wrote the warning a year ago. The warning has become the data. The data points to an institution that is teaching its young officers, in real time, that the silent path is the survivable path.

The institution survives on the opposite habit. It survives on the officer who tells the boss the plan will not work. The Pentagon Hegseth has built selects against that officer. The Pentagon of 2046 will be staffed by the people that selection produces.

The Pin

Warrior ethos carries two definitions inside the same Pentagon. The stated definition: a combat-experienced professional, disciplined, ready, lethal. The operational definition: an officer whose judgment matches the political layer's. The first definition includes Donahue. The second excludes him.

The boundary test, when you watch this from outside: when a leader's stated standard and operational standard point in different directions, the operational one is the real definition. The stated one is doing the PR work.

The script for your own room, when you hear a magnet word doing heavy work — readiness, alignment, performance, fit, accountability:

"When you say [the word], do you mean the published definition, or the one we are actually selecting for? Because I think those are two different things in this room."

That sentence is twenty seconds. It costs nothing. It surfaces the drift before it becomes selection. It is what an adult institution sounds like, asked one word at a time, by the officer who has not yet decided whether the silent path is worth the survival it promises.

That officer is who future-us is waiting on.

Subscribe now

Jerry W. Washington, Ed.D., is a retired Marine Corps Master Sergeant (23 years, MOS 1371 Combat Engineer, 1993–2016), USC Rossier School of Education graduate, Instructor at UCI Division of Continuing Education, and independent researcher. He writes What Time Binds on Substack.

Share

Share

A few days ago I posted a piece about General Daniel “Chappie” James, the Black four-star general whose portrait was removed from the Pentagon’s Air Force Art Gallery this spring. The post went wider than I expected. About 40,000 reactions, 10,300 comments, 19,000 shares. Most of the comments split along the lines you would expect. A smaller number tried to think out loud. One of those was a real conversation.

The commenter said he did not believe in any form of DEI, and that Chappie James deserved his place on that wall. I asked him what he meant by DEI before we kept going. He came back with a careful answer. DEI, he said, is a broad term not defined in Title VII. It has been interpreted many ways. He distinguished it from Affirmative Action with quotas for minorities (which we will see is incorrect). Its purpose (DEI), he said, was to ensure qualified minorities get the same chance as other qualified candidates.

I addressed most of what he said. I pointed out that his definition of DEI’s purpose actually answers his own opening question. I noted the body of law that organizational DEI policies actually cite. I asked whether his use of the word “minorities” included veterans, working mothers, older workers, and people with disabilities, since those are the largest groups protected under the same legal infrastructure.

I missed one word. Quota.

What follows is what I should have said.

My colleague published a companion piece on The Unstated this week, working the same mechanism on a different word from the same thread. He pinned “merit.” This essay pins “quota.” Both words carried weight that was not defined in the comments.

Subscribe now

Why “quota” is a magnet word

A magnet word is a term that pulls multiple meanings into the same conversation. Different people use it. The word stays constant. The meanings underneath shift. When everyone is using the same word for different things, the conversation looks like it is making progress while drifting further apart.

“Quota” pulls toward at least two meanings.

The first is the legal definition. A quota is a fixed reserved number of positions for a specific demographic category. UC Davis Medical School’s program in the mid-1970s, which reserved 16 of every 100 admission slots for minority applicants, is the textbook example. The Supreme Court ruled that arrangement unconstitutional in 1978. Quotas in that sense have not been lawful in the United States for forty-seven years.

The second is the political usage. In this register, “quota” expands to cover any numeric target, any preference based on race or sex, any goal that mentions demographics, and any outcome where the racial or gender composition of a workplace changes. The political usage has detached from the legal meaning. People use it to describe things the law has never required and things the courts have specifically prohibited, often in the same sentence.

When someone says “Affirmative Action with quotas for minorities,” they may mean either of these. Until you ask, you do not know. The conversation cannot move while both definitions are loose in the room.

What the record shows

The phrase “affirmative action” first appeared in federal policy in 1961. Executive Order 10925, signed by President John F. Kennedy on March 6, 1961, required federal contractors to “take affirmative action to ensure that applicants are employed, and that employees are treated during employment, without regard to their race, creed, color, or national origin.” The order created the President’s Committee on Equal Employment Opportunity, chaired by Vice President Lyndon Johnson. The text says nothing about quotas. The instruction was to actively not discriminate.

In 1965, President Johnson signed Executive Order 11246, which replaced 10925 and put real teeth on the requirement. Federal contractors had to take affirmative action on race, color, religion, sex, and national origin. The Office of Federal Contract Compliance Programs was created to enforce it. Compliance work included non-discrimination policies, outreach, recruitment, and demographic tracking. No quotas.

The first time the federal government used numeric targets was the Philadelphia Plan, developed in 1967 and implemented by the Nixon administration in 1969. Federal construction contractors in Philadelphia had to set numeric goals and timetables for minority hiring. The plan used goals, which differ from quotas in a way that matters legally and operationally. A goal is a target you aim for and document good-faith efforts toward. Missing it does not violate the law as long as you can show the effort. A quota is a fixed reserved number that must be filled regardless of qualified applicants. The Philadelphia Plan worked in goals.

Then came the case that defined what affirmative action could and could not do. In 1973 and 1974, Allan Bakke, a white engineer in his thirties, applied twice to UC Davis Medical School and was rejected both times. The medical school ran a special admissions program that reserved 16 of every 100 seats for minority applicants. Bakke sued. The case rose to the Supreme Court.

On June 28, 1978, the Court issued its ruling in Regents of the University of California v. Bakke. It was a 5-4 decision with six separate opinions. Justice Lewis Powell wrote the controlling opinion. The Court ruled that the UC Davis quota system was unconstitutional.

Race could be considered as one factor in admissions decisions where multiple factors are weighed together. Reserved seats for racial categories violated the Equal Protection Clause of the 14th Amendment and the Civil Rights Act of 1964.

Bakke is the case that banned quotas. The popular shorthand treats Bakke as the case that established affirmative action. It gets that backward. Bakke is the ceiling on what affirmative action can do. It has been the ceiling for forty-seven years.

The Court revisited the question several times after. In 2003, Grutter v. Bollinger upheld race as one factor among many in law school admissions at the University of Michigan. The same day, Gratz v. Bollinger struck down Michigan’s undergrad point system, which gave automatic points based on race. The pattern held. Race as one factor among many was permitted. Race as automatic points or reserved seats was not. Fisher v. University of Texas upheld the same approach in 2016. In 2023, Students for Fair Admissions v. Harvard ended race-conscious admissions in higher education entirely.

In January 2025, President Trump signed Executive Order 14173, which revoked EO 11246 and ended the federal contractor affirmative action mandate for race, color, religion, sex, sexual orientation, gender identity, and national origin. The grace period ended April 21, 2025.

That is the record. Federal affirmative action ran from 1961 to 2025. It was built on non-discrimination, outreach, and goals. One university medical school used a quota for five years. The Supreme Court ruled it unconstitutional in 1978. No federal contractor was ever required to use a quota. The Court reaffirmed the quota ban every time the question came back. Higher education ended race-conscious admissions in 2023. The federal contractor mandate ended in 2025.

The word “quota” describes a thing that was tried at a small number of institutions for a short period and was ruled unconstitutional half a century ago.

Why I didn’t address it the first time

Two reasons.

The first is that the commenter’s broader frame was partially correct. DEI is broader than affirmative action. His definition of DEI’s purpose, ensuring qualified people get the same evaluation based on their record, is accurate. I led with the agreement because that is the right opening. You start where you actually agree, and you give the other person credit for what they got right.

The second is that I underestimated how much weight the word “quota” was carrying in his frame. It came in as a parenthetical inside a longer sentence. I read it too quickly. The word was doing structural work. It told me what he thought affirmative action was. The word was carrying nearly five decades of political language that the legal record cannot bear. I should have flagged it.

I am flagging it now.

Pin the term

In law, a quota is a fixed reserved number of positions for a specific demographic category. The Supreme Court ruled this unconstitutional in Bakke in 1978. It has remained unconstitutional in every case since.

The definition includes:

• UC Davis Medical School’s 16-of-100 reserved seats (1973 to 1978, struck down in 1978)

• Any hypothetical rule that reserves a fixed number of positions for a specific demographic category

• Any “hire one X candidate for every Y candidates of another category” requirement

The definition excludes:

• Affirmative action plans that document outreach efforts

• Race-conscious admissions that weighed multiple factors together (legal 1978 to 2023, ended by Students for Fair Admissions)

• Numeric goals and timetables (legal under EO 11246, ended by EO 14173)

• Recruitment outreach to underrepresented groups

• Reasonable accommodation processes under the ADA or Title VII religious accommodation rules

• Demographic data tracking through EEO-1 reports

When someone uses the word “quota,” ask which of these they mean. If they describe Bakke-era reserved seats, agree with them. Those were unconstitutional and have been since 1978. If they describe something else, name what they are actually describing and use the right word for it.

The one-minute script

The next time you hear “Affirmative Action with quotas,” try this:

When you say “quota,” do you mean a fixed reserved number for a demographic category? That has been unconstitutional since 1978. The Bakke case banned it. If you mean recruitment outreach, multi-factor admissions, or numeric goals, those are different practices with different legal histories. Which one are you talking about?

The point is to ask what the word means before the conversation keeps going. Once the word is pinned, the conversation has somewhere to go.

Log it

Words outlive the things they point to. “Quota” was a real policy at a small number of programs in the 1970s. The Supreme Court ended it in 1978. The word kept living in political speeches, voter messaging, opinion commentary, and family conversations for forty-seven years. People on both sides of the affirmative action debate have argued about quotas long after the courts settled the question.

That is how political language drifts.

The term computes the same on every screen. It carries weight in every conversation. It points at nothing the law has allowed since the Carter administration.

The repair is to ask what the word means and to write down what each person meant by it. The next conversation starts cleaner because this one was pinned. The thread about Chappie James reached 39,000 people. The conversation about the word reached one. That second conversation is where the work actually happens.

The Bakke decision was June 28, 1978. Carter was president. The Bee Gees had four songs in the Billboard top ten. The Sex Pistols had broken up four months before. That is how long this word has been doing political work the law cannot back.

When you hear it, ask what it means. Write down the answer. The record needs the receipt.

Jovanny Suriel’s analysis of “merit” from the same Facebook thread is at The Unstated. Read it alongside this one.

Subscribe now

Thanks for reading What Time Binds. Subscribe for free to receive new posts.

Jerry W. Washington, Ed.D., is a Marine combat engineer veteran, independent researcher, and instructor at UCI Division of Continuing Education. He writes What Time Binds on Substack.

Ten of thirteen executives say they are “very clear” or “clear” on what the company will look like after the change. Eight of thirteen say the leadership team is “aligned.” Then the consultants ask each executive to write the specific ways the new company will be different.

Three of the written answers describe three different companies.

One executive writes: the system will be larger and more complex, but operations processes will be very similar to present. Another writes: the change is mostly about standing and competing alone without a parent corporation in charge. A third writes: there will be new assets, new markets, different cost structure, new people, and new leaders.

Same CEO. Same change program. Three companies on paper.

That scene comes from “The False Alignment Trap,” a piece by Julia Dhar, Kristy Ellmer, and Philip Jameson in the July–August 2026 issue of Harvard Business Review. The reporting is sharp. The word doing the damage is aligned.

Background on the Piece

Dhar leads BCG’s People and Organization Practice in North America. Ellmer leads BCG Transform. Jameson is an associate director at the firm. The article previews a forthcoming book, How Change Really Works, from HBR Press in 2026. The central claim is that most organizational change efforts fail. Historically, 50 to 70 percent of them have, and that figure has not improved in three decades. The failures often trace to executive teams behaving as if they agree on the change when they do not.

I have been holding this piece since it ran in May. Sent it to my co-researcher, , the same week. Both of us considered its points carefully, came back to it twice, talked it through. It is the kind of HBR article that gets the diagnosis right, and the work behind it is serious. Coordination failure under pressure is a problem more researchers are naming directly now. That is good news for the field. It is also good news for the people inside the meetings where the failure happens. The diagnosis is the easy part.

Stay with this entry through to the end. There is a Monday WTB essay coming on June 15. The close tells you what it will do and why today’s field guide is the setup for it.

Why “Aligned” is a Magnet Word

“Aligned” sounds like a status report. It feels descriptive. It functions as a placeholder.

People use the word aligned to end meetings the same way they use it to start change programs. It signals that no further conversation is required. The signal is false more often than it is true.

Dhar and her co-authors put it plainly:

When leaders say “we are aligned,” what they usually mean is “we are not in one another’s way.” Sometimes they mean “we have discussed this topic at least once and generally accept the contours of a plan.” Those meanings sound exactly like a third meaning — “we have agreed on the specifics, the owners, and the timeline” — that very few teams have actually reached.

That is the magnet. The same word pulls four different agreements into the same sentence, and unless you are aware and have the tools to check, the drift happens.

What People Actually Mean When They Say “Aligned”

Here are the meanings the word collects.

1. Out-of-the-way alignment. “We are not blocking each other.” Lowest bar. Often reached just by avoiding open conflict in the room.

2. Discussed-once alignment. “We talked about it. The contours look okay.” No specific commitments. No trade-offs named. No owners assigned.

3. Conceptual alignment. This was the COO’s word in the HBR piece. One executive said “I agree with the first proposal but not aligned with the second.” A second said “I am aligned.” A third said “I am partly aligned.” The COO closed the meeting with “I think it is clear that we are conceptually aligned.” Nobody was. Nobody said so.

4. Specific alignment. “We agree on the decision, the trade-offs, the owners, and the timeline. We can each write the same four sentences without checking with each other first.”

The first three meanings sound like the fourth. They are not the fourth.

The Failure Mode

When the four meanings collide, each leader walks back to their team with a different operational version of the same word. The change program splinters by Friday. Each business unit pursues a slightly different version of the same transformation, all under the same banner. More drift.

Dhar and her co-authors describe three downstream patterns: paralysis (lots of talk, no action), hyperactivity (lots of action on the wrong things), and tunnel vision (steady progress on a narrow read of the goal). All three share one upstream cause. The word “aligned” ended a meeting before the agreement had been pinned.

Pin It

In this room, aligned means we have each written down the specific decision, the named owner, and the deadline — and we can each produce the same four sentences without consulting each other. (Think about a more detailed RACI diagram).

Includes:

• A specific commitment with a verb and a date.

• A named owner for each piece.

• A written trade-off (what we are not doing as a consequence of doing this).

• A scheduled date to revisit.

Excludes:

• “Conceptually aligned.”

• “I think we’re speaking the same language.”

• “I’m on board with the spirit of this.”

• A head-nod followed by silence.

Boundary Test: Pandora

Alexander Lacik, the CEO who led Pandora from 2019 to 2025, ran the test in his first weeks on the job. The management team had forty-six priorities. Lacik took the team off-site for two days and said the group would not leave until twelve priorities remained. The owner of each priority defended it in front of the room. The team voted.

Another executive team in the HBR piece took a different route. Anyone who said the word “aligned” in a meeting paid five dollars. The fines funded a celebration when the team hit a specific goal.

That is what teams do when they have figured out that “aligned” lies until proven otherwise.

One-Minute Script (Use Monday)

Before any meeting closes on “we’re aligned,” ask four questions out loud:

• “What specifically did each of us just agree to do?”

• “What does done look like, and by when?”

• “Who owns each piece?”

• “What are we not doing as a consequence?”

Write the four sentences. Get them visible to everyone in the room before the meeting ends. If the sentences cannot be written, the word does not apply.

Log It

Open a definition and decision log. Date it. Write one line:

“In this room, today, ‘aligned’ means a written decision, an owner, a deadline, and a named trade-off. We will revisit on [date].”

Sign it. Send it. The log is what future-us inherits when this week’s meeting becomes next quarter’s blame conversation.

Carry It Into Your Week

One rule: no meeting closes on “aligned” until four sentences are written and visible.

If the team cannot produce the sentences, the team is not aligned. They have discussed the topic. That is a different thing, and it deserves a different word.

On Monday, June 15, the WTB essay picks up where this entry leaves off. The HBR piece names the symptom: leaders behaving as if they agree when they do not. It does not say what happens to meaning below the C-suite after the CEO broadcasts the slogan and the change program moves into the org. That is where the harder work lives.

Monday’s piece treats meaning repair as installed practice across the org. The HBR five-step process happens once, at the top, during a transformation. Installed practice happens every Tuesday afternoon at 2 PM, in every meeting where someone says “we’re aligned” and everyone stops to truly get aligned.

If today’s field guide gives you the practice for the next meeting, Monday’s gives you the architecture for the year. Please share and follow!

Share

Subscribe now

Thanks for reading What Time Binds. Subscribe for free to receive new posts.

Jerry W. Washington, Ed.D., is a Marine combat engineer veteran, independent researcher, and instructor at UCI Division of Continuing Education. He writes What Time Binds on Substack.

Share

Someone tells you they use AI. You say, Okay. You have just learned almost nothing. Did the machine do the work, or did it stand next to them while they did it? Those are opposite arrangements. There is one small word that covers both.

Monday, I took the team-scale fight over the word “AI,” master it against kill it, and showed how naming the use dissolves the whole argument. Today, the word is smaller and closer. Use. The verb you reach for when someone asks how AI fits into your day.

“I use AI” pulls toward three different jobs, and they do not belong in the same sentence.

The rep. The task where the doing is the point. Writing the hard paragraph. Working the problem until it cracks. The output matters less than what producing it builds in you.

The chore. The task where the doing teaches nothing. Reformatting a table. Cleaning up a transcript. Hand it over, and you lose nothing worth keeping.

The counterfeit. A rep wearing the chore’s clothes. It feels like busywork, so you hand it off without a thought, and it quietly takes a capability you needed. You notice nothing, because the output looks fine.

Subscribe now

What goes wrong

Trouble starts when you treat all three as one “use.” You hand off reps believing they are chores. The work still ships, so the dashboard says you are doing great. Meanwhile the thing that made you good at the work drains out, one handoff at a time.

I spent a long review on this a few years back, on what generative tools do to a writer’s belief in their own ability. The short version: ability is built by doing hard things and finding out you can. Skip the doing, and you keep the proof while you lose the thing it was proving.

The sorting question

Before you hand any task to the machine, run one question.

If the tool vanished tomorrow, could I still do this, and would I know if I did it badly?

That sorts the three.

Yes to both. It is a chore or a skill you already own. Hand it off. You are buying time. The ability was already yours.

The doing is how you would build the skill in the first place. That is a rep. Keep it. Let the tool spar with you, react to your rough draft, push back. Keep your hand on the first sentence.

You cannot answer the second half, because you would not know good work from bad. That is the counterfeit, and it is the one to watch. Not knowing whether you could judge the output is the signal that you are about to hand away the judgment.

Watch it land

Three ordinary moments.

The email sitting in your drafts. Most email is a chore. The reply that confirms a time, the note that forwards a file. Hand those over and save the minutes. The hard one is different. The apology. The pushback to your boss. The message where the wording is the relationship. There, the writing is the thinking. Draft that one yourself.

The paragraph you cannot start. The blank screen is the rep. Learning to start is a skill, and you build it only by sitting in the discomfort until a bad sentence comes. Hand the opening to the machine, and you will need it again tomorrow, and the day after, because you never built the muscle. Let the tool argue with your bad sentence. Keep the bad sentence yours.

The homework your kid swears they finished. This is the counterfeit in its purest form. The assignment was the rep, the struggle that builds the mind that can do the next one. The kid used the tool, kept the grade, and lost the learning. Here is the trap: the kid cannot tell. The judgment that would catch the gap is exactly the judgment the shortcut skipped. Chieng’s friend speedran Buddhism and did not reach enlightenment. Your kid speedran the worksheet and cannot tell that anything is missing.

The one-minute install

Say it out loud before you hand something over.

To yourself: Is the doing the point here? If the tool disappeared, could I still do this, and would I know if I botched it?

To your kid: Did the AI do it, or did it help you do it? Can you do the next one without it?

That second question is the whole game. “Can you do the next one without it” is the line between a tool and a crutch.

Log it

Keep a short list. Your recurring tasks, sorted into rep, chore, and counterfeit. Read it again every few months, because the line moves. As the tool gets better, today’s honest chore can become tomorrow’s counterfeit, good enough now to hide the loss. The stronger the machine gets, the more carefully you have to name the use.

Questions for the week

What did you hand to AI because it felt like a chore, and was it?

What could you do a year ago that you would now reach for the tool to finish?

If the tool went dark tomorrow, which of your “uses” would expose a skill that quietly drained?

If you are raising or teaching someone, what rep is the assignment actually for, underneath the answer it asks for?

What we hand forward

Chieng’s whole speech came down to one line: the work makes the person. A tool that removes the work removes the formation. That holds at your desk, and it holds at your kid’s.

What we pass to the people coming after us is not the output. We pass them, whether we kept the reps that build judgment or counterfeited our way through, leaving them a world full of confident people who cannot grade their own work.

Name the use. Keep the reps that matter. Give the chores away without guilt. Stay in the driver’s seat, and teach the next driver to do the same.

Subscribe now

Jerry W. Washington, Ed.D., is a retired Marine, a UCI instructor, and the publisher of What Time Binds. Monday’s companion piece, on how this same cut works at team scale, is here. The research on AI and a writer’s belief in their own ability is in “The Impact of Generative AI on Writer’s Self-Efficacy” (SSRN 4538043, 2023).

Share

Monday’s essay ended with Pope Leo XIV refusing the comfortable theological floor of dignity and walking toward the labor floors. He gestured at contribution and consent, but left procedural alone. The encyclical was theological in form and labor doctrine in content.

Three floors still need walking. The contribution floor needs the chain the Pope only pointed at. The consent floor needs the institutional-versus-relational pin that the encyclical implies but does not hold. The procedural floor never got a Vatican stage Monday at all. This is the field guide entry that walks them.

The “floors” come from Richard Feynman. Asked in a famous interview why magnets attract, he refused the surface question and asked back: what do you mean by “why”? Every why-question rests on a floor, an assumption the listener has agreed to stop questioning. Like a market, justice, God, history, and lived experience floor. The floor depends on the listener, and the same word means different things depending on which floor it lands on. Monday’s essay used Feynman’s coceptual framing to set up the four floors of dignity. Wednesday walks the three the Pope did not take.

I had planned to write this alone. However, the week handed me three writers instead. Arimitsu published a piece called “Layers of Sameness” that maps the five different things people can mean when they say two words refer to the same thing. Chris Stephens reshared an April 30 essay this morning on LinkedIn that extends my earlier “accountability” piece into structural territory I had not taken it. Eric Patterson sent me a Hollis Robbins essay on what “higher” means in higher education. The Robbins piece is for my Friday Bindings. The other two are working below. As you can tell, I love to collaborate on knowledge…because that’s What Time Binds: knowledge.

Subscribe now

The diagnostic Arimitsu handed me

Arimitsu’s argument, compressed: when two people say the same word in a meeting, they may be measuring sameness by five different criteria, each one operating like a different floor of the same building.

Category-same. Two Tokarev pistols. Different factories, different metallurgy, different machinists, same blueprint. The word stays steady because it points at the design.

Descent-same. Theravāda Buddhism and Zen. Both descend from the same source. The lineage holds the name. What gets done in any given room varies widely.

Collectively-agreed-same. The Kusanagi sword at Atsuta Shrine. Almost no one has seen it. It counts as real because a community agreed to treat it as real.

Recognition-anchored-same. The ship of Theseus. The name was attached at a moment of public recognition. The boards have all been replaced. Whether the object still has any right to the name depends on how far you have drifted from the moment of attachment.

Relationship-same. The merchant knows the ship is a forgery. The boy who saved up to buy it treasures it as authentic. Same object. Two different verdicts. Both internally coherent. Both measuring by criteria the other isn’t using.

A magnet word like dignity sits on top of all five layers. The four floors I described Monday are typologies of where the word lands. The five layers are typologies of which kind of sameness the speaker is claiming when they use it. Floors plus layers is the real coordinate. A pin in the room needs both. The complexity stays, and the comprehension improves.

The procedural floor

Procedural dignity is the floor where people stand when they say “we will treat you fairly because we have a fair process.” The pin is the process. The dignity is in following the procedure, applying it consistently, giving everyone access to the same set of steps.

The default layer here is collectively-agreed-same. We all said this is the procedure. Recognition-anchored-same operates underneath: the procedure was set up at a moment when the community said, “this is the way.” Both work as long as nobody asks how the procedure got built and who was there when it did.

Chris Stephens named the specific failure mode this floor produces. She reshared the diagnosis on LinkedIn this morning. The pattern is “accountability without authority.” Organizations routinely hold people accountable for outcomes they never controlled. The procedure assigns responsibility, which is procedural dignity, doing something else. It also withholds the authority required to discharge that responsibility. The procedure says you matter. The structure says you do not get to act on what matters.

Same word, opposite function. An organization built around repair and harm reduction uses “accountability” to maintain coherence. An organization built around profitability uses the same word to manage optics and locate blame. The procedural floor honors a word both kinds of organizations use. Most people cannot tell which one they are using until the consequences arrive.

The failure mode in Arimitsu’s terms: descent-same gets erased. The procedure looks the same. Where it came from, what it was originally for, and who it was designed to protect have all dropped out. The category-same of “we have a process” runs over the descent-same of “this process came from this history, with these original includes and excludes.” Procedure without lineage is procedure that cannot tell whether it is doing the work it was built to do.

The contribution floor

Contribution dignity is the floor where people stand when they say “you have earned a place in this.” The pin is the contribution. The dignity tracks what you have built, made, fixed, raised, taught, served, or fought for.

The default layer is category-same. You are in the category of people who contribute. Relationship-same operates underneath in particular cases: your specific work in this team, this family, this unit. Both layers work when contribution is visible and recognized.

The Pope pointed at this floor Monday. He did not walk it. He named the AI labor supply chain (content moderators, miners, data labelers) and said their contribution is what makes the visible AI possible. The gesture was clear. The pin still needs writing.

The failure mode this floor produces, in Arimitsu’s terms: descent-same gets erased. The category-same of “we recognize contributors” runs over the descent-same of “the contributions that built the floor everyone is standing on.” The recognized contributor is the visible one, in the current cycle, doing the legible work. The contributions that came before, the contributions that made the visible work possible, the contributions that got recoded as “support” or “service” or “inputs” all drop out.

The clearest current example is in higher education, which is what Hollis Robbins walks in a piece Eric Patterson sent me this week. The work that builds the visible contributor depended on contributions that the recognition system never coded as contribution at all. Same shape of failure, different room. Robbins gets the Friday Binding treatment.

The consent floor

Consent dignity is the floor where people stand when they say “you cannot be acted on without your agreement.” The pin is consent. The dignity tracks whether the people in the action gave the action permission.

The default layer is relationship-same. Consent happens between specific parties. The merchant and the buyer. The doctor and the patient. The participant and the researcher. Descent-same operates underneath: the modern human rights tradition is the lineage that built the current understanding of what consent requires.

The Pope gestured at this floor Monday in the language of new forms of slavery. Workers in the AI supply chain cannot refuse the conditions of their labor. That is one shape of consent failure. The AI training data question is another. The people whose words, faces, voices, code, and art trained the current generation of models were not asked. The models were trained on their work as if the work were available.

The failure mode in Arimitsu’s terms: collectively-agreed-same overrides relationship-same. An institution claims consent on behalf of people who never gave it. The relationship between the labeler and the lab, the artist and the model, the writer and the training set, never happened in the form of consent. The institution writes a policy that says consent was obtained because the data was publicly available, or because terms of service were technically agreed to, or because the use claimed protection under fair use doctrine. The institutional pin overrides the relational one.

This is the structural mismatch Chris Stephens named in her LinkedIn post, applied to a different room. Responsibility without authority is one face of it. Consent without agreement is another. Both are pins that the institution writes on behalf of someone who cannot push back.

The full pin

A What Do You Mean (WDYM) pin on a magnet word needs both coordinates.

Floor: Which kind of dignity is being claimed in this room? Theological, procedural, contribution, consent.

Layer: Which kind of sameness is being claimed when the word gets used? Category, descent, collectively-agreed, recognition-anchored, relationship.

The full pin reads: “In this room, dignity means [floor-specific definition], measured by [layer of sameness], with these includes, these excludes, and a date to revisit.” Without both coordinates, the room is operating on a partial coordinate, and the drift will start immediately.

Scripts

For the procedural floor: “When you say we will hold them accountable, what authority are they being given to act on what they are accountable for? If we cannot name the authority, the accountability is decorative.”

For the contribution floor: “When you say this person earned their place, whose unrecognized work made the place possible? If we cannot name that chain, the recognition is operating on a partial map.” (Think of the term, standing on the shoulders of giants).

For the consent floor: “When you say consent was given, who gave it, and could they have refused without losing what they were trying to do? If the answer is no, the consent is not consent.”

Friday tee

Eric Patterson sent me a piece by Hollis Robbins this week called “What does the ‘higher’ in higher ed mean?” Robbins is reading the word higher the way I have been reading dignity, the way Chris Stephens is reading accountability, the way Arimitsu is reading same. Same shape of failure. Different magnet word. I am holding it for Friday.

Return

Three writers in conversation across one week, and a Pope in Rome refusing the comfortable floor, and the field guide entry is provided.

The word is dignity. The floors are procedural, contribution, consent. The layers are five. The pin is both. The room is wherever you are.

Future-us is reading.

Previously, on What Time Binds: The Word Is Dignity walked Monday’s encyclical and the apology no pope had made before. Feynman’s Real Question Was “What Do You Mean?” developed the floors concept this piece builds on. What Do You Mean, “Crime Against Humanity”? ran the UN slavery vote through the same diagnostic.

Which floor is your room standing on without knowing it? Tell me in the comments.

Subscribe to What Time Binds for Friday’s Binding, the monthly Saturday Pinned Terms strip, and next week’s Monday essay.

Subscribe now

Share

In 1876, the editors of Princeton’s student newspaper argued against exam proctors. Watching students for fraud was, they wrote, “a means of bad moral education.” Treat a person as presumptively dishonest, and some will oblige you; treat the same person as honorable, and they learn to behave that way. Their proposal: every student signs a pledge that they received no help, and the faculty find better work than patrolling the room.

That proposal became the Princeton Honor Code in 1893. It was modified only lightly across the next 133 years. It outlasted two world wars, the 1960s, Watergate, search engines, and SparkNotes.

This month, the Princeton faculty voted to put the proctors back.

The Honor Code is technically still in place. Students still sign the pledge. Now, a professor stands at the front of the room to confirm it is true. Rose Horowitch reported the vote in The Atlantic on May 12. The system that ran without watchers now has watchers.

I teach, and I have watched this pressure arrive in real classrooms, so the phrase is worth pinning before it loses the last of its shape.

Subscribe now

The word

“What do you mean by ‘the honor system’?”

You say it at work more than you think. Expense reports. Remote hours. Open-book policies. Self-certification. Self-checkout. The phrase shows up wherever an institution decides not to check.

Why it is a magnet

The phrase pulls three ways, and the three do not agree with each other.

One: trust without verification. Nobody watches, and you behave anyway.

Two: a social code held up by peers. You sign a pledge. You are expected to report the people who break it.

Three: a compliance label on a watched process. The pledge stays stapled to the top. The words “honor system” remain in the handbook.

Princeton’s faculty meant the first. The code ran on the second. AI creation forced the third. The phrase never changed through any of it. That is the magnet at work: one set of words holding three meanings, with everyone certain they share one.

The failure mode

An honor system holds only while the cost of breaking it stays high enough that most people choose not to. Generative AI took that cost to almost nothing. A unique essay, in any style, in seconds, with typos added so it reads as human.

The collapse then arrives all at once, at a tipping point. Nadia Makuc, a Princeton senior and former chair of the Honor Committee, put the mechanism in one line: visible cheating breeds more cheating. The numbers track it. The Committee on Discipline found 82 students responsible for violations in 2024–25, up from 50 in 2021–22. In a senior survey of 501 students, 30 percent said they cheated, 28 percent used a chatbot on an assignment where it was barred, and 45 percent knew of a peer cheating and said nothing. The students who follow the rule start to feel like the only ones still holding it.

The tool did not cheat anyone. It changed the price of cheating, and once the price fell the rest followed — visible violations, the belief that everyone was doing it, and an institution that answered by putting watchers back. The honor system did not die of generative AI. It died of what generative AI made cheap and what Princeton did next.

Pin it

In this room, “the honor system” means a coordination rule that holds only while three conditions are true at the same time:

1. The cost of breaking it stays real.

2. Breaking it stays visible.

3. Most people believe most people comply.

Lose any one of the three and what is left is an unenforced rule wearing the old name.

Boundary test

Includes: expense self-reporting, take-home exams, remote-work hours, open-source attribution, self-checkout.

Excludes: anything already proctored, audited, logged, or version-tracked.

Example: Princeton, 1893 to 2026.

Non-example: a proctored final with a signed pledge stapled on top. The pledge is theater once the proctor is in the room.

The one-minute script

“When you say we run on the honor system here, which part do you mean? That we trust people, or that nobody checks? Those two came apart somewhere, and I want to know which one we are relying on before I sign.”

Log it

Write down which of the three conditions your team actually depends on. Date it. Revisit it when the cost of breaking the rule drops, because something always drops it. At Princeton the something was generative AI. In your shop it will be something else, and it will arrive without an announcement.

What gets passed forward

The 1876 editors saw part of this. They warned that suspicion manufactures the rogue it claims to catch. They were right, and being right did not save the code.

The word got passed forward intact. The conditions that made it mean something stayed behind in a different century.

When you inherit a word, check whether you also inherited the conditions that made it work. “Honor system” still sounds like 1893. It now operates like a proctored room with a signature page. Same words. Different machine underneath.

Ask the question before you sign the pledge. What do you mean by the honor system?

Share

Subscribe now

What Do You Mean? is a weekly field guide inside What Time Binds. One magnet word, pinned and made usable, every Wednesday.

— Jerry Washington, Ed.D. • what-time-binds.com

Subscribe now

On February 14, 2025, Peter Kyle stood at the Munich Security Conference and announced that the United Kingdom’s AI Safety Institute would be renamed the AI Security Institute. Same building, staff, and budget. The mandate narrowed: chemical and biological weapon misuse, cyberattacks, and child sexual abuse material. Out of scope: bias, freedom of speech.

Four months later, the United States did its version. The Commerce Department renamed its AI Safety Institute the Center for AI Standards and Innovation. Howard Lutnick said the rebrand removed “censorship and regulations” used “under the guise of national security.” NIST guidance instructed CAISI partners to drop references to “AI safety,” “responsible AI,” and “AI fairness.”

The buildings did not change, but the contents did.

If you watched the Bletchley Declaration in November 2023, the Seoul Summit in May 2024, and the Paris AI Action Summit in February 2025 in sequence, you watched the phrase “AI safety” walk into a meeting room, sit down, and discover it had been holding two conversations the whole time. At Paris, fifty-eight countries signed a Statement on Inclusive and Sustainable AI. The United States and the United Kingdom refused. JD Vance, on the dais: “The AI future is not going to be won by hand-wringing about safety.“

That moment is diagnostic. The word everyone had been using at every summit for two years did not survive: it drifted.

A Magnet Word

“AI safety” is one symptom. The bigger word is “AI governance.”

Some words pull people into different interpretations while feeling like agreement. They keep meetings moving, people signing contracts, and policies passing. Then, six months or six years later, someone opens the deliverable and discovers that the room had been using the same syllables to describe four different jobs.

Watch what happens when you pull on “AI governance.”

In Brussels, it means the European Commission’s coordinating function across twenty-seven member states under the AI Act. In Washington, until January 2025, it meant Executive Order 14110; after January 2025, it meant Executive Order 14179, which rescinded 14110 in three days. In Sacramento, it meant SB 1047 (until Governor Newsom vetoed it in September 2024); then it meant SB 53, which kept transparency reporting and dropped the kill-switch. In Denver, under the Colorado AI Act, it means a developer’s duty of “reasonable care” to prevent algorithmic discrimination. In an Anthropic Responsible Scaling Policy, it means an internal Risk Officer, a capability threshold, and a Long-Term Benefit Trust. In a Deloitte board roadmap, it means a skills matrix and a refresh schedule. At AI Now, it means antitrust and worker organizing. At UNESCO, it means human rights and dignity grounded in 194-state agreement.

These are different objects. They share a vocabulary and almost nothing else.

What I Watched For Twenty-Three Years

I spent twenty-three years inside institutions where ambiguous language carried operational consequence. Marine Corps construction sites. Logistics commands across four countries. Corporate facilities portfolios in three states. A 23-office closure managed across HR, Legal, Finance, IT, and Business Development through five years and a pandemic. The pattern showed up every time. The words looked clear. The room nodded. The cost arrived later.

In 2005, I took over the facilities portfolio for 3D Marine Logistics Group in Okinawa: 300 facilities, three million square feet, 30 locations across Japan, Korea, Hawaii, and Guam, 4,600 personnel. I reported to a Chief of Staff who used the word “ready” daily. Every commander in the chain used “ready” daily. Nobody stopped to verify that “ready” meant the same operational state to the building maintenance officer, the procurement officer, and the host-nation liaison. We coordinated as if we agreed. When we didn’t, projects slipped.

I left active duty in 2016, finished a doctorate at USC Rossier in Organizational Change and Leadership in 2023, and spent the next two years on a scoping review of 131 sources across eight disciplines: healthcare communication, aviation crew resource management, cognitive science, military operations, organizational psychology, linguistics, high-reliability organizing, and team science. The literature confirmed what I had watched in every room I had ever worked in. The default state of human communication is drift. Shared understanding takes work. The work is to catch the drift before the budget or the policy pays for it.

Eight Rooms, Eight Words

Eight communities use “AI governance.” They mean different things.

Regulators in Brussels mean rights protection through ex ante conformity assessment and member-state market surveillance. Regulators in Washington, under the current administration, mean removing barriers to American AI dominance.

Frontier-AI labs (Anthropic, OpenAI, Google DeepMind) mean an internal Responsible Scaling Policy or Frontier Safety Framework: their own capability thresholds, their own auditors, their own stopping rules.

Big Four consultants mean a board-level oversight discipline producing model inventories, third-party risk management, and ISO 42001 mappings against the NIST AI Risk Management Framework.

Alignment researchers (Bengio, Hinton, Russell, the twenty-five-author Science paper from May 2024) mean binding licensing, mandatory pre-deployment testing, and one-third of AI R&D budgets devoted to safety. Closer in shape to nuclear-style controls than to NIST-style voluntary frameworks.

Civil-society scholars (Buolamwini, Gebru, Bender, Whittaker, Crawford, Benjamin) mean redistribution of power. Antitrust. Worker organizing. Bans on weaponized facial recognition. Refusal of the surveillance business model. Sarah Myers West, in AI Now’s 2024 paper Power and Governance in the Age of AI: “AI as we know it today is a creation of concentrated industry power.”

DoD and federal contractors mean a Chief AI Officer role under OMB M-24-10, an inventory of “rights-impacting” and “safety-impacting” AI systems, and compliance with the five Responsible AI principles (Responsible, Equitable, Traceable, Reliable, Governable).

UNESCO and the African Union mean development capability, digital sovereignty, and human rights anchored in 194-state agreement. From outside the Brussels-Washington axis, “AI governance” often reads as a Western export imposed on jurisdictions that did not draft it.

Same word. Eight rooms. Eight non-overlapping objects.

Where Drift Turns Into Cost

The fault lines are where decisions actually turn.

Catastrophic-risk control versus algorithmic-discrimination prevention. Bengio, Hinton, GovAI, METR, and the Future of Life Institute treat governance as preventing extinction-class harms from frontier systems. Buolamwini, Gebru, the Algorithmic Justice League, and the Colorado AI Act treat governance as preventing harm to people who are already being injured. The two camps publicly disagree about which deserves the word.

Voluntary risk management versus enforceable regulation. NIST AI RMF, the Hiroshima Code of Conduct, the OECD Principles, and most corporate Responsible Scaling Policies are voluntary. The EU AI Act, the Colorado AI Act, NYC Local Law 144, and SB 1047 (briefly) carried teeth. The EU version comes with fines up to 7 percent of worldwide turnover. The voluntary versions come with press releases.

Innovation enablement versus power redistribution. Executive Order 14179, signed January 23, 2025, defines governance as removing barriers to “America’s global AI dominance.” AI Now’s 2025 Artificial Power report defines governance as breaking up tech oligarchy. Both call themselves AI governance. Both cite the same underlying technology. They share no policy primitives.

A vote on SB 1047 turned on which definition was operative. The Paris no-signs turned on which definition was operative. The CAISI rebrand turned on which definition was operative. When real consequences attach to a magnet word, somebody’s definition wins, and somebody’s loses, and the people whose definition lost rarely got to vote on which one was on the ballot.

Four Questions to Run in Any Room

Here is the test I run in any room where the phrase “AI governance” is used.

• Whose harm are we governing? The public’s, the company’s, the state’s, or humanity-at-large?

• Who pays if we are wrong? Users, taxpayers, shareholders, or future people?

• Who enforces? A regulator, a board, an auditor, or a market?

• Who decides what “wrong” means? Courts, voters, engineers, or executives?

If two people in the room give different answers to any of these questions, you are looking at a magnet word event. Stop the meeting. Pin the term.

The one-minute script:

“When we say ‘AI governance’ in this conversation, are we talking about (a) preventing catastrophic harm from frontier models, (b) preventing discrimination in consequential decisions, (c) ensuring board-level oversight of enterprise AI use, or (d) protecting national competitiveness? I want to make sure we are working from the same definition before we agree on next steps.”

That sentence costs sixty seconds. It saves whatever the consequence would have cost on the back end.

Governance Worth the Name

Governance worth the name names four things.

• It names the harm: who gets hurt, in what way, on what timeline, with what evidence.

• It names the actor: which institution, which person, which budget line is responsible for the decision that creates the harm.

• It names the enforcement: which regulator, court, auditor, or contracting officer can stop the harm or reverse it.

• It names the redress: what the injured party gets when the harm happens, who pays for it, and how soon.

Anything that calls itself governance while omitting any of these four is, in Emily Bender’s phrase, marketing.

Under that test, NYC Local Law 144 is governance: narrow, contested, and badly enforced (the New York State Comptroller’s December 2025 audit found 75 percent of complaints misrouted), but the four parts are present. Anthropic’s Responsible Scaling Policy, version 3.0, fails the test. It names the harm and the actor. It does not name the enforcement (Anthropic audits Anthropic) or the redress (no injured party has standing). The same is true of OpenAI’s Preparedness Framework and Google DeepMind’s Frontier Safety Framework. These are commitments. Calling them governance softens the word until it cannot do the work.

What This Means

The most precise users of the term “AI governance” right now are working from outside the rooms where it is operationalized. Bender and Hanna’s The AI Con (May 2025). Whittaker’s NDSS keynote. Buolamwini’s Algorithmic Justice League. AI Now’s Artificial Power report. They name the harms. They name the actors. They describe what enforcement and redress would look like. Do they get invited to write the policies?

The companies that draft Responsible Scaling Policies fund the academic centers that evaluate them. Consulting firms that sell governance services write the frameworks regulators cite. The U.S. and U.K. security services rebrand institutes without legislative input. The word “AI governance” sits in the middle of all of it, doing the work of agreement while the underlying definitions move in opposite directions.

If you are in a meeting next week where someone uses the phrase, ask which version they mean. If they cannot answer in one sentence, you are in a vocabulary problem pretending to be a governance conversation. Somebody is going to pay for that mistake. The only question is whether you saw it coming.

Subscribe now

Jerry W. Washington, Ed.D., is a twenty-three-year U.S. Marine Corps retired Master Sergeant. Ed.D., USC Rossier (Organizational Change and Leadership) and author of Simulated Realities: Generative AI and the Remanufacture of Professionalism (2023) and a 131-source scoping review on shared meaning under pressure (working paper, 2026). He’s also an independent advisor on AI readiness for education and workforce systems: K–12 districts, community colleges, workforce boards, and veteran-service organizations. He publishes weekly at What Time Binds. Engagements and writing at jerrywwashington.com.

Subscribe now

I started the week sketching a Pinned Terms strip about accountability with my daughter.

The setup wrote itself. Juno Tanaka stands at the whiteboard. Six people in the Ops Room. The card on the wall reads ACCOUNTABLE. Each character has a different definition loaded in their head, and none of them know it. Juno reaches up, pulls the card down, holds it in two fingers like contraband, and says, “This one’s a trench coat.”

That’s the gag. Three panels. A magnet word doing what magnet words do, a character calling it out, a small repair move written on the wall. Comic done.

Then I read Kendra Okposo’s HBR piece, “Accountability Must Be Chosen, Not Mandated,” and I closed the laptop on the strip.

The strip would still work as a strip. The problem was scope. What three panels can show needs forty minutes in a real meeting, and Okposo had already done most of the work to set that meeting up. Her case study with a global transportation provider is honest reporting. Her Mindset/Meaning/Mechanisms structure is clean. Her core finding earns the column inches HBR gave it: you cannot impose accountability, you can only build the conditions where people choose it.

I want to push on one of her three Ms: Meaning.

What Okposo means by Meaning

Okposo defines Meaning as the why that makes the choice worth it. Values like integrity. A shared goal that transcends any one person’s contribution. Care and commitment that survive when scrutiny fades.

That definition is correct, and it is doing only half the job.

There is another version of Meaning operating in the same conference room, and it breaks teams long before motivation gets a chance to do its work: the referent of the word accountability itself. Not why people should choose it. What they think they are choosing.

In a business performance meeting, six leaders agree they need to be more accountable. One means I will own the result regardless of what my function controlled. Another means I will surface bad news earlier. Another means I will stop blaming other functions in front of leadership. Another means I will deliver what I committed to or call it out before the deadline. Another means I will let myself be measured against numbers I disagree with. Another means I will do what I am told without asking why.

Six votes for accountability. Six different votes. The motivational why does not fix this, because each person is choosing a different thing. The room sounds aligned, but it isn’t.

This is the meaning failure that runs underneath the meaning failure Okposo names.

The case study is the proof

Read her case study carefully.

The transportation company’s leaders worked with BTS to “co-author a playbook” that broke accountability into “real, recognizable moments in their work, both great and not great.” Okposo lists the moments they pinned. Name the real problem under the symptom. Claim your role in the solution. Raise enterprise-level concerns even when they cross your function’s lines.

That co-authoring is the move. The leaders were not motivated into accountability. They wrote a shared definition of the word and then practiced it across three ten-week cycles. Thirty weeks of repetition mattered. So did the sequencing. The definition came first.

Okposo files this under Mindset and reinforces it under Mechanisms. Read it again, and the pinning sits in the middle: a working agreement on what accountable points to, in this company, in these meetings, this quarter. Without that agreement, the mindset shift has nothing to anchor to, and the mechanisms enforce six different things at once.

She did the work. I would name it Meaning, in the sense of shared reference, and call it the deeper repair move her clients actually executed.

Pin it in the room

If you sit on a leadership team where accountability keeps surfacing as a problem, run a forty-minute session before the next quarterly review. Pull the word off the wall. Treat it as a magnet word until the room has decided otherwise.

Three questions, written down, agreed by everyone present.

What does accountability include here? Two or three concrete behaviors. Surfacing a missed milestone within forty-eight hours. Naming what you personally controlled in the outcome. Bringing your function’s blockers to the all-hands before someone else has to.

What does it exclude here? Two or three behaviors people sometimes confuse with it. Taking blame for outcomes you did not influence. Public self-criticism that signals humility without changing behavior. Quiet compliance with a target you privately believe is wrong.

When do we revisit? Name the date. Quarterly is honest. Annual is a slow walk into the same drift.

The artifact is the agreement. It lives somewhere future-us can find it. The next leader who joins the team reads it before their first business performance review. The next time someone says “we need more accountability around here,” the room can answer: we have a working definition. Are we executing it, or do we need to revise it?

That is the conversation Okposo’s transportation company actually had. The playbook is the artifact. The thirty weeks were the practice. The Mindset shift was the result.

Back to the strip

I might still draw the comic. Juno’s trench coat line is too good to leave on the cutting room floor.

A strip can show the collision. It cannot carry the sentence the team writes after. That sentence is what holds the team together in the next quarter, the next reorganization, the next time the word accountability shows up on a slide, and everyone nods at a different meaning.

Choose accountability, by all means. First, decide together what you are choosing.

A few questions for your team

1. The last time someone on your leadership team said “we need to be more accountable,” what did they mean? What did the rest of the room hear?

2. If you wrote down the includes/excludes for accountability in your organization right now, would the document agree with the behaviors your performance reviews actually reward?

3. Who on your team has been carrying blame for outcomes they did not influence, because no one paused to define what accountability meant for that work?

Subscribe now

Jerry W. Washington, Ed.D., is a retired Marine Corps Master Sergeant, a UCI Division of Continuing Education instructor, and the publisher of What Time Binds. He writes about meaning, time-binding, and how teams lose and recover shared reality under pressure.

If this piece was useful, forward it to one person on a leadership team that keeps having the same accountability conversation. Then subscribe so the next one finds you.

Subscribe now

I researched this essay on the day it happened.

March 25, 2026. One hundred and twenty-three nations voted to declare the transatlantic slave trade the gravest crime against humanity. Three voted no. The United States was one of them.

I pulled the resolution text. I read the U.S. explanation of the vote. I read the EU’s abstention statement, the UK’s statement, and France’s separate legal filing. I mapped the magnet words, ran the vote through my framework, and started writing.

Then I closed the document.

I closed it because my hands were shaking, and I have learned, the hard way, over years of doing this work in public, to tell the difference between writing with conviction and writing with heat. I had heat. I could feel it in the first paragraph. In the way I was reaching for words that accused rather than words that explained.

So I waited. I told myself I would come back when I could do what I ask my readers to do: name my priors, hold the strongest counterargument, and let the evidence lead.

A month later, I came back. The heat is still there. I am a Black man. I am the descendant of enslaved people. I spent 23 years in the Marine Corps building infrastructure for a country whose founding documents excluded my ancestors from the category of “person.” When my country votes no on a resolution about slavery, my body has an opinion before my brain arrives.

That is my prior. I carry it. I name it here because I believe in doing the work I ask of others: examining the maps I inherited before I use them to read new terrain.

Here is where the frustration sharpens. I know my priors. I have spent years interrogating them publicly in essays, in a research framework, and in a scoping review of 131 academic sources across eight disciplines. I wrote an entire essay called “Race as a Ghost Map” that opens by admitting the emotional weight I bring to these questions and then lays out a self-audit protocol anyone can use.

The question that kept me from publishing in March, and brings me back now, is whether the people who cast that “no” vote, and the 52 nations who abstained, have done the same work on their own maps. Because the evidence from the General Assembly floor suggests they have not.

The magnet word at the center

On March 25, 2026, the UN General Assembly adopted Resolution A/80/L.48 by a vote of 123 in favor, 3 against, and 52 abstentions. Ghana introduced the resolution on behalf of all 54 African Union member states. The United States, Israel, and Argentina were the three “no” votes. Every EU member abstained as a bloc. The resolution declared the trafficking and chattel enslavement of Africans “the gravest crime against humanity” and called for reparatory justice.

The magnet word is “crime.”

For the 123 nations voting yes, “crime against humanity” carries a specific moral and legal weight. It means: this happened, it was wrong by any standard human beings have ever held, the consequences persist in measurable ways, and the international community has an obligation to address those consequences. The word “gravest” adds emphasis. The word “reparations” names the obligation.

For the three nations voting no, and for much of the abstaining bloc, the same phrase triggers a different reading entirely. “Crime against humanity” sounds like a legal instrument being applied retroactively. “Gravest” sounds like a ranking that diminishes other atrocities. “Reparations” sounds like a financial claim without a clear recipient.

Same words. Different pictures. And because nobody stopped to say, “When we say crime, do we mean a moral declaration or a legal indictment?” the entire vote became a collision of unstated definitions.

This is what I study. This is what my Meaning Repair as Cognitive Infrastructure framework maps. And this UN vote is that framework running at full civic scale.

The four phases, live from the General Assembly

Drift. The meaning of “crime against humanity” has been drifting for 25 years. The 2001 Durban Declaration called slavery “a crime against humanity and should always have been so,” a carefully crafted phrase that acknowledged the moral reality while sidestepping retroactive legal application. That formulation was already a compromise. It contained two competing definitions inside a single sentence: slavery is a crime (present tense, moral), and it should always have been (subjunctive, acknowledging it was not legally designated as such at the time). For a quarter century, that ambiguity held. Both sides could point to the same words and see their own meaning reflected back. Drift felt like agreement.

Suppression. Four consecutive U.S. administrations boycotted Durban follow-up events: Bush, Obama, Trump, Biden. Every boycott is a suppression move at institutional scale. Drift was detected. The mismatch between the moral reading and the legal reading was known. And the response, each time, was to leave the room. The research on suppression in my scoping review describes this pattern precisely: silence spreads through storytelling. Each administration inherited the precedent of the last administration’s exit, and the bar for re-entry rose. By the time Ghana introduced this resolution in March 2026, the U.S. had 25 years of accumulated suppression infrastructure, inherited silence dressed as diplomatic continuity.

I wrote about this mechanism in “Version Control for Democracy.” At the President’s House in Philadelphia, NPS employees removed interpretive panels about enslaved people from a jointly curated historic site without notice or consultation. Judge Rufe’s ruling used version-control language to describe the damage: public memory had been edited by power without process. Her order was a restore command.

The UN vote is the same pattern at global scale. For 25 years, the Western bloc has been editing the shared record by subtraction: leaving conferences, boycotting commemorations, abstaining from votes. Each exit removes a panel from the interpretive exhibit. The simplified story fills the gap on its own.

Repair Activation. Ghana’s resolution is itself a repair move. Foreign Minister Ablakwa walked to the General Assembly podium and, in effect, said: “We have been using the same words and meaning different things for a quarter century. Here is what we mean. Here are six specific components: apology, restitution, compensation, rehabilitation, satisfaction, and guarantees of non-repetition. Here is the legal standard we are invoking: jus cogens. We are pinning this term.”

That is term pinning. It costs more than 20 seconds at this scale, and the mechanism is identical to what happens in a meeting room when someone finally says, “When we say finalize, do we mean the spec is locked, or the date is confirmed?”

Outcome. The pin held for 123 nations. For the rest, it failed. And now future-us inherits two competing artifacts: a General Assembly declaration endorsed by a supermajority, and a formal U.S. explanation of vote that reads like a counter-definition. The drift between these two maps is wider now than before the vote, because the vote forced the incompatible definitions into the open. Successful repair in part of the room, failed repair in the rest. The feedback loop is running.

A coalition of scholars and reparations advocates issued a warning in the weeks following the vote that names the risk precisely: the resolution could stall at declaration, with reparations language diluted into development assistance, reduced to symbolic gestures, and deferred indefinitely through what they called “death by commission.” In framework terms, that is the Phase 4 danger: the pin held on the day of the vote, and the definition can still drift back if the institutional follow-through stalls.

My priors, on the table

Here is where I need to slow down and do the thing I ask readers to do.

I carry priors into this analysis. I am the descendant of people who were property under American law. My family history contains gaps that will never be filled because enslaved people were recorded as inventory, not as persons. When Ambassador Negrea said the resolution’s supporters were engaged in “cynical usage of historical wrongs as a leverage point in an attempt to reallocate modern resources to people and nations who are distantly related to the historical victims,” my body read that sentence before my mind parsed it. The phrase “distantly related” landed in a place where reason does not live first.

So I will name what I notice in myself:

I notice that I want the U.S. to be wrong. That is a prior, and I hold it.

I notice that I feel the word “distantly” as erasure, as if the distance between me and my great-great-great-grandmother, whose name I will never know because no one recorded it, is a matter of diplomatic convenience. That is an emotional response, and it carries real information; it is a signal, and I treat it as one.

I notice that when Negrea said, “President Trump has done more for Black Americans than any other president,” I stopped reading his argument and started reading his credibility. That is a suppression risk in my own processing. If I dismiss the entire U.S. position because one sentence offended me, I am doing the same thing I critique: letting the emotional charge of a magnet word shut down the grounding check.

So I went back and read the full explanation of vote. I read it as if it were a team member’s position in a high-stakes meeting where I disagreed with the conclusion and needed to understand the reasoning.

The legal core of the U.S. argument rests on rejecting retroactive application of international law. The position holds that acts from the 15th through 19th centuries cannot constitute violations of jus cogens as that term is understood in contemporary international law. This is a real argument. It has legal weight. Scholars disagree about it, and reasonable people can land on different sides.

The argument about ranking atrocities, that calling slavery the “gravest” crime, objectively diminishes the suffering of countless victims and survivors of other atrocities,” is also a real concern. Israel shares it. The EU shares it. The anxiety about creating a hierarchy of suffering is genuine, even if I believe it misreads the resolution’s intent.

I can hold these things. I can acknowledge the legal argument while believing, with everything in my body and my training, that the moral answer is clear and that the legal objection functions, in practice, as a mechanism for avoiding the moral conclusion.

The thing that breaks the repair loop is the refusal to examine one’s own priors with the same rigor.

The frustration: unexamined maps on the Assembly floor

In my “Race as a Ghost Map” essay, I wrote that shaming people rarely updates maps. Ghost Maps thrive in defensiveness. If we want different outcomes, we have to get better at the craft of updating.

I still believe that. And I have tried to live it: auditing my own assumptions publicly, inviting readers to do the same, building a framework grounded in evidence rather than accusation.

This vote tested my patience with that posture. Because what I saw on the Assembly floor was a room full of nations with deeply unexamined maps, maps inherited from the same era that produced the slave trade itself, defending those maps with procedural language while 123 nations asked them to do one thing: say the full sentence.

The EU abstained because the resolution used superlatives. France warned against “pitting historical tragedies against each other.” The UK said slavery must “never be forgotten” while insisting there is no legal duty to address it. These are institutional suppression moves. Drift detected. Repair declined. Room exited.

The U.S. explanation of vote went further. It deployed a “distantly related” frame that erases the living, measurable consequences of slavery: wealth gaps, health disparities, incarceration rates, educational access, as if 400 years of compounding disadvantage disappears when you add enough generations to the denominator. And it wrapped a partisan claim about presidential records inside a diplomatic statement to the General Assembly.

Ambassador Negrea’s office did not interrogate its own priors. It performed them.

And there is a deflection gaining traction in the month since the vote that deserves attention: the claim that “Africans sold Africans,” offered as if it dissolves the case for reparations. This is another magnet word at work. “Participation” is collapsing the distance between selling war captives into a regional system that predated European contact and building a four-century industrial apparatus of chattel enslavement across three continents, capitalized by state-chartered corporations, enforced by national navies, and codified into property law. As one analysis published this month in the Mail & Guardian put it, no African kingdom built the legal infrastructure of chattel slavery or created the plantation economy and the financial instruments that sustained it. The “Africans sold Africans” argument functions the same way “distantly related” functions: it lowers the resolution on the map until the uncomfortable structural detail disappears.

That is the distinction I keep coming back to. I am willing to say: here is my map, here is where it might be wrong, here is the emotional weight I bring, and here is my evidence. The frustration, the real, bone-deep frustration, is directed at institutions and actors who refuse to do the same work. Who treat their inherited position as the ground state and every challenge to it as an intrusion.

In Ghost Map terms, that refusal has a name: self-exemption drift. The belief that your map requires no audit because it arrived before you did. Everyone else has priors. You have principles.

The democracy connection

This vote did not happen in a vacuum. It landed inside a measurable democratic recession.

I have written about this pattern across multiple essays, in “Greenland, Venezuela, and the Time-Binding Collapse,” in the Reclaiming Dr. King piece, in “Version Control for Democracy.” The through-line is consistent: when democratic institutions weaken, the space for meaning repair shrinks. Autocratizing systems compress the room for the question “What do you mean?” because that question, by its nature, distributes power. It gives the person hearing a word the right to ask whether the person speaking it meant what they think it means. That is a democratic act. It requires the assumption that both parties in the exchange have standing.

The UN vote maps onto this pattern. The 123 nations that voted yes were performing a democratic function at the international level: we disagree about this term, so we will define it, vote on it, and record the result. The three nations that voted no, and the 52 that abstained, were performing a different function: we will decline to participate in the definition process while retaining the power to act as if our definition is the default.

That asymmetry is what autocratization looks like in meaning systems. The powerful party does not need to win the argument. It needs to leave the room and continue operating under the old map as if the vote never happened.

In my time-binding language, this is space-binding behavior: the mindset of territorial control in the present moment, at the expense of the intergenerational record. The resolution asked the international community to bind time, to declare that what happened between the 15th and 19th centuries carries obligations into the present and future. The “no” vote said that time is over, and its claims expired with it. The Korzybskian Gap, between our capacity to transmit knowledge across generations and our willingness to act on what that knowledge demands, is right here, on full display, in a General Assembly chamber with 193 seats.

What would repair look like?

I am a meaning repair researcher. I do not get to diagnose the drift and then walk away. That would make me guilty of the same suppression I am critiquing.

So here is my attempt at pinning the terms in this room:

“Crime against humanity,” in this context, means: a moral designation that carries present-day obligations. It includes the recognition that the consequences of the transatlantic slave trade continue to structure economic, health, and social outcomes for people of African descent worldwide. It excludes the claim that legal instruments must be applied retroactively to create criminal liability for individuals. It includes the obligation of institutional accountability. Boundary test: if a nation accepts that slavery was a “crime against humanity” and refuses any form of reparatory engagement, the moral designation is decorative.

“Reparations,” in this context, means: a structured, multi-component response that may include formal apology, institutional reform, educational initiatives, restitution of cultural property, economic investment, and policy change. It includes truth-telling. It includes the acknowledgment that “we have addressed this” requires evidence, not assertion. It excludes the reduction of all reparatory claims to cash transfers from one government to another. Boundary test: if the only version of reparations you are willing to discuss is the one you find easiest to reject, you are not discussing reparations.

“Gravest,” in this context, means: a description of scale, duration, and structural consequence, supported by the historical record of a system that operated across four centuries, three continents, and tens of millions of human lives, with institutional legacies that remain measurable. It excludes the claim that other atrocities are diminished by this designation. Boundary test: if your objection to calling slavery the “gravest” crime is that it might minimize the Holocaust, ask yourself whether you raised the same objection when the Holocaust was described in equivalent terms. If you did not, the concern is selective.

The sentence I am trying to say

I am offended by the vote. I have earned the right to be, and I am not going to pretend otherwise.

I am also trained enough, and honest enough with myself, to know that my offense is a signal, and signals inform decisions; they do not make them. It tells me where my identity is fused to the question. It tells me where I need to slow down and check my own maps. And I have done that, here, in public, the way I always try to.

What I am asking, from the U.S., from the EU, from every nation that abstained, is the same thing I ask from teams in my workshops, the same thing I teach in my course, the same thing the MRCI framework installs as a daily practice:

Say what you mean. Then check whether the person across the table heard what you intended.

Because “crime against humanity” is wearing a trench coat. It is carrying at least three different definitions into every room it enters: a moral acknowledgment, a legal designation, and a call to action. The 123 nations that voted yes tried to pin all three. The U.S. rejected all three. And the 52 who abstained declined to choose.

That is meaning failure at civilizational scale. And meaning failure, left unrepaired, compounds. It always does.

I know my priors. I have named them. I have subjected them to evidence, to counterargument, to the discomfort of admitting where my emotional response might outrun my analysis.

The question I carry out of this vote is the same one I carry out of every MRCI analysis: are you willing to do the same? Or does your map arrive pre-audited, inherited, defended, and exempt from the scrutiny you demand of everyone else’s?

Because if your answer is the second one, the drift will keep compounding. And future-us will inherit a map with the hard parts subtracted, the bolt holes still visible where the panels used to hang, and no restore point in sight.

There is a date on the horizon that makes this concrete. September 2026 marks 100 years since the 1926 League of Nations Slavery Convention, the first international treaty to ban slavery. The African Union has designated 2026 through 2036 as the Decade of Action on Reparations. Ghana’s strategy is visible now: secure the General Assembly declaration first, then use it as institutional backing for binding mechanisms. The two competing artifacts I described, the 123-nation declaration and the U.S. counter-definition, will collide again, in a more specific venue, on a shorter timeline. Future-us is already arriving.

The fix, as always, starts with four words.

What do you mean?

Subscribe now

Previously in this series: Version Control for Democracy explored the same meaning-repair pattern at the President’s House in Philadelphia. Race as a Ghost Map examined the cognitive machinery that keeps outdated maps of human difference in circulation. Greenland, Venezuela, and the Time-Binding Collapse mapped the democratic recession and the Korzybskian Gap.

The MRCI framework is introduced in full in The Research Behind What Time Binds. The term “self-exemption drift” was coined in the What Time Binds lexicon. “Trench coat,” Juno Tanaka’s term for a magnet word hiding multiple meanings, comes from Pinned Terms.

A What Do You Mean? essay — the What Time Binds series, where we examine the words doing more work than most people think in our civic life.

I wrote a version of this essay seventeen days ago, on the day the executive order dropped.

Thursday, March 26, 2026. I read the White House fact sheet. I read the full text of Executive Order 14398. I opened a document and started writing. I got about three thousand words in: the compliance officer in Northern Virginia, the career fair question, the definition that swallows itself. I thought I had the story.

Then I stopped. Something felt incomplete, maybe because the order was new and the legal analysis was thin. It seemed I was reacting to the headline, and I have spent enough time studying how words work under pressure to know that reacting to headlines is how you end up fighting ghosts.

So I sat on it. I told myself I would come back in two weeks and see what the professionals were saying: the employment attorneys, the compliance firms, the people whose job it is to tell companies what government language actually requires.

I came back this week.

What I found was worse than what I wrote 2 weeks ago.

Subscribe now

The question that starts everything

Here is the scene I kept ruminating on.

Thursday afternoon, March 26, 2026. A compliance officer at a defense contractor in Northern Virginia opens her laptop. There is a new executive order, signed that morning. She reads the title: “Addressing DEI Discrimination by Federal Contractors.” She reads the mandatory contract clause her company must adopt by April 25 (at the time, that was 30 days away). She reads it again.

Then she picks up the phone and calls her general counsel.

“We sponsor a table at the Howard University career fair every spring,” she says. “Is that still legal?”

Her general counsel pauses. “I need to read the definition section again.”

They are both educated people. They know what “discrimination” means. They have spent years making sure their company follows federal anti-discrimination law. The problem is that, as of that Thursday, they cannot tell whether their compliance program is now in violation.

That pause, that uncertainty, is where the story usually stops. The headline says “DEI ban.” People take sides. Supporters say it is about restoring merit, and critics say it is about rolling back civil rights. Everyone argues about the politics.

I want to take you somewhere else. I want to show you the machinery underneath, the part that does not make the headlines but does the most damage. Because there are two words buried in the contract clause that changes everything about how this order works. And it seems I am the only one talking about them.

The first word is “discrimination,” but it is not the word you think

Since the Civil Rights Act of 1964, “discrimination” has carried a stable meaning in federal contracting. It meant: treating individuals differently based on race, color, religion, sex, or national origin, in ways that deny them equal opportunity.

Under that definition, companies built diversity programs as tools to prevent discrimination. Executive Order 11246, signed by Lyndon Johnson on September 24, 1965, required federal contractors to take active steps to address employment barriers. The logic was direct: if your workforce looks nothing like the available labor pool, and you cannot explain why, something in your process is filtering people out. Find it. Fix it.

That order stood for sixty years. On January 21, 2025 (Trumps first full day in office), President Trump revoked it through Executive Order 14173, “Ending Illegal Discrimination and Restoring Merit-Based Opportunity.”

The new executive order — EO 14398, signed fifteen months later on March 26, 2026, finishes the job. It defines “racially discriminatory DEI activities” as:

Read “allocation or deployment of an entity’s resources” slowly. A mentoring program for first-generation professionals? An employee resource group? A career fair table at a historically Black university? As the law firm Lexology noted, the order “includes an expanded definition of prohibited DEI activities that could create risk around programs previously considered lawful.”

The word “discrimination” has been turned around. It now points in the opposite direction from where it pointed for six decades. The compliance officer cannot tell exactly where the new line is.

That vagueness matters. But it is the starting point, not the ending point. Because what I found when I came back to this story is that the vagueness is connected to something much more specific and much more dangerous.

The second word is the one that actually matters

When I read the law firm analyses this week, I expected to find arguments about the definition. I did. Every firm flags the vagueness. Every firm warns that the definition of “racially discriminatory DEI activities” is broader than existing civil rights law.

But the word that kept showing up in bold, in red, in capital letters, across Skadden, Arnold & Porter, Winston & Strawn, Jackson Lewis, DLA Piper, and a dozen others, was not “discrimination.”

It was “material.”

Here is the sentence in the mandatory contract clause, the one every federal contractor must sign by Friday, April 25, 2026:

“The contractor recognizes that compliance with the requirements of this clause are material to the Government’s payment decisions for purposes of section 3729(b)(4) of title 31, United States Code (False Claims Act).”

If you are not a lawyer, that sentence looks like boilerplate. It reads like the fine print nobody reads. It sounds like it belongs in the same category as “I have read and agree to the terms of service.”

It is the most consequential sentence in the entire order.

Here is what it means, in plain language: every time your company sends an invoice to the federal government (every progress payment, every milestone billing, every monthly statement), you are now certifying that you are not running any program that could be called a “racially discriminatory DEI activity.” If you are running such a program, and you submit that invoice anyway, you may have just committed fraud against the United States government.

That is what “material” does. It connects a vague definition of discrimination to a very specific, very old, very powerful law. As Winston & Strawn warned, this language “is designed to eliminate one of the most common defenses in FCA cases, the argument that the violated requirement was not material to the government’s decision to pay.”

A 160-year-old law with a bounty attached

The False Claims Act was signed by Abraham Lincoln on March 2, 1863. Its original purpose was to stop defense contractors from selling the Union Army sick mules and defective gunpowder. If a contractor lied to the government about what they were delivering, they owed triple damages.

The law still works essentially the same way. If a contractor knowingly submits a false claim for payment, they owe three times the government’s damages, plus penalties per claim. It is the government’s primary tool for fighting contractor fraud, and it has recovered hundreds of billions of dollars over its lifetime.

But here is the part that changes the game: the False Claims Act has something called a “qui tam” provision. Qui tam comes from a Latin phrase meaning “he who sues on behalf of the king.” In modern English, it means this:

Any private citizen can file a lawsuit on behalf of the federal government alleging that a contractor submitted a false claim. If the case succeeds, that citizen collects between fifteen and thirty percent of the recovery.

Any citizen. A coworker. A former employee. A competitor. Someone who reads your company’s website and notices that you have a supplier diversity program.

The potential payouts are substantial. In large government contracting cases, whistleblower shares can reach into the millions.

And the Department of Justice is actively looking for these cases. At the Federal Bar Association’s Qui Tam Conference in February 2026, a month before this executive order was signed, Deputy Assistant Attorney General Brenna Jenny, who leads the Commercial Litigation Branch in the DOJ’s Civil Division, said her office is prioritizing anti-discrimination False Claims Act cases for expedited review. She said whistleblower reports are already triggering investigations across automotive, defense, pharmaceutical, technology, and telecommunications industries.

In May 2025, the DOJ had already launched its “Civil Rights Fraud Initiative”, explicitly aiming to “utilize the False Claims Act to investigate and, as appropriate, pursue claims against any recipient of federal funds that knowingly violates federal civil rights laws.”

Law firms are already advertising for clients. One firm's website has a page titled "DEI Whistleblower Rewards," telling potential whistleblowers that they may be sitting on "one of the most valuable whistleblower claims in America." Another practice that has litigated three False Claims Act cases to the Supreme Court and recovered over a billion dollars for whistleblower clients lists DEI compliance as a top area of increased enforcement scrutiny in 2026, noting that any organization that certifies compliance to the government while operating otherwise "may face FCA exposure."

Let me say that plainly. The chain works like this:

A vague definition of “discrimination,” one broad enough to cover a career fair table or a mentoring program, gets attached to a mandatory certification clause. The certification clause contains the word “material,” which links compliance to every invoice. Every invoice becomes a potential false claim. And a Civil War-era bounty system pays private citizens to find and report the violations.

Your coworker can now get paid for reporting your company’s HBCU recruitment partnership as government fraud.

How the machine runs without anyone pulling the lever

Here is what I think most people miss about this order, regardless of where they stand on the politics.

The order does not need to be enforced to work.

It works through what I think of as a chain of reasonable fears. Each link in the chain makes perfect sense to the person holding it. Nobody is being irrational. Nobody is panicking. Everyone is doing the math.

The compliance officer’s math: I have a program that could fit the broad definition. My general counsel cannot tell me with certainty that it does not. The definition will not get clearer before the April 25 deadline. If I keep the program and we get audited, my company faces contract termination, debarment, and triple damages under the False Claims Act. If I shut the program down, the cost is one less career fair table.

She shuts it down.

The general counsel’s math: I have twenty programs that touch race in some way, mentoring, recruitment outreach, employee resource groups, and supplier diversity. I can mount a legal defense for each one. But defending even one False Claims Act investigation will cost six figures in legal fees before we get to a ruling. And the definition is vague enough that I cannot guarantee we win. The safer advice is to pause everything and wait for clarity that may never come.

He advises pausing.

The subcontractor’s math: My prime contractor is now required to report me if they know or “reasonably should know” that I am violating the clause. They are going to ask me about my programs. If I have a supplier diversity initiative, I need to explain it in writing to a company that has every incentive to flag me rather than risk their own contract. The safest move is to dismantle my program before anyone asks.

She dismantles it.

The employee’s math: I just read on a law firm’s website that I could collect fifteen to thirty percent of the recovery if I report my company’s DEI programs under the False Claims Act. My company has a Black employee resource group that gets conference room space and a small budget. Is that “allocation or deployment of resources based on race”? I do not know. But the law firm says I should call for a free consultation.

He calls.

Every person in this chain is making a rational decision based on the information available to them. Nobody needs to be racist. Nobody needs to be malicious. The machine runs on reasonable fear, and reasonable fear runs on vague definitions.

This is what I mean when I say the uncertainty is the enforcement mechanism. The order does not need a single prosecution to achieve its goal. It needs the math to work. And the math works right now, today, thirteen days before the first deadline.

The strongest argument for this order and where it breaks

I owe this section to anyone reading who thinks the order is necessary. I have spent years arguing that you have to take the strongest version of the opposing position seriously, and I am not going to stop because this one is personal.

The core claim: some DEI programs crossed the line. Some organizations set demographic targets that functioned as quotas. Some evaluated employees partly on identity characteristics rather than qualifications. Some training programs treated people differently based on their race in ways that created hostile environments. When that happens, it is discrimination under Title VII, and it was illegal before this executive order existed.

That claim has evidence. The Supreme Court found on June 29, 2023, in Students for Fair Admissions v. Harvard, that Harvard’s admissions process violated the Equal Protection Clause. Some corporate DEI programs have been challenged in court for creating hostile environments. The principle that individual merit should drive employment decisions has broad public support across demographic groups. I have discussed this before here.

I take that seriously.

And the executive order still fails on its own terms. Here is why.

If the goal were to eliminate race-based quotas and preference systems that violate Title VII, the order could say that. It could define the prohibited conduct with enough precision that a compliance officer could draw a clear line. It could create a process for getting a binding interpretation. It could distinguish between programs that create racial preferences (already illegal) and programs that expand access to opportunities, legal, effective, and standard practice for decades.

The order does none of this. As Lexology’s analysis noted, “certain activities that may previously have been viewed as compliant with civil rights laws could now be risky” under the expanded definition. Arnold & Porter flagged that the order’s own legal authority, the Federal Property and Administrative Services Act, “faces headwinds” in multiple circuit courts that have read the statute narrowly and may not support regulating contractors’ internal operations at all.

The definition is wide enough to cover mentoring programs, recruitment outreach, and supplier diversity initiatives, activities with documented business value and no legal vulnerability under existing civil rights law. The vagueness forces contractors into a choice: over-comply by dismantling effective programs, or accept risk exposure by maintaining programs that might be flagged.

When you add the False Claims Act bounty to that choice, the math tilts hard toward dismantling. A company that keeps its HBCU partnership faces potential triple damages and a whistleblower lawsuit. A company that shuts it down faces... nothing.

The strongest argument for the order is about eliminating illegal preferences. The order as written eliminates the space for legal, effective, evidence-based talent development. Those are different projects. The vague definition lets them look the same. The word “material” makes the consequences identical.

What this looks like in practice, April 25 and beyond

The first deadline hits in thirteen days. By Friday, April 25, every federal agency must insert the new clause into all contracts, with flow-down to subcontractors at every tier.

By Sunday, May 25, the Federal Acquisition Regulatory Council must issue interim guidance, meaning the clause must be in contracts a full month before the guidance explaining how to interpret it arrives. Winston & Strawn called this “a notable timing gap” that “may create implementation uncertainty in the near term.”

By Friday, July 24, every agency head must report on compliance to the Assistant to the President for Domestic Policy. The Office of Management and Budget will identify “high-risk sectors” for additional scrutiny.

And the DOJ has been directed to ensure “prompt review” of whistleblower lawsuits with a decision on whether to join each case within sixty days.

The practical effect, already visible across the law firm analyses: companies are not waiting for enforcement. They are conducting what attorneys call “DEI health checks,” privileged reviews designed to identify any program that could be characterized as a “racially discriminatory DEI activity” under the broadest possible reading. The advice, nearly universal across the firms I read, is to assume the broadest reading will apply.

Several attorneys noted something else worth paying attention to. The executive order specifically excludes sex-based programs from its definition; it targets only race and ethnicity. And enforcement authority is decentralized across individual contracting agencies rather than a single specialized office. That means different agencies may interpret the same clause differently. A program that passes muster at NASA might draw scrutiny at the Department of Defense. The compliance landscape is vague and variable.

What I want you to notice

I have written this piece without my usual framework language. No four-phase models. No Latin terms for communication breakdowns. No citations to my scoping review.

I did that on purpose, because I want anyone reading this, regardless of whether you have ever heard of my work, regardless of your politics, regardless of whether you think DEI programs are essential or excessive, to see the machinery.

Here is what I want you to notice:

The word “discrimination” changed direction, and the sentence it lives in did not change shape. That is how meaning capture works. The container looks the same. The contents are opposite. And the people reading the sentence cannot tell, in the moment, that anything moved.

The word “material” did more work than any other word in the entire executive order. It connected a policy position to a fraud statute. It converted a compliance question into a financial weapon. And it activated a bounty system that turns private citizens into enforcers. Most people who have opinions about this order have never read the sentence it lives in.

The vagueness is the product. A precise order would prohibit specific conduct and allow everything else. A vague order prohibits a category so broad that the safest response is to eliminate everything in it. The compliance officer cannot tell whether her career fair table is legal. That uncertainty is what makes her shut it down. A clear rule would not produce that result. The ambiguity is doing the work.

The bounty changes the incentive structure for everyone. Before this order, a coworker who disagreed with your company’s mentoring program could complain to HR. Now, that coworker can file a qui tam lawsuit and potentially collect millions. That does not mean every coworker will. It means every person who runs a program knows that every coworker could. The possibility reshapes behavior before a single suit is filed.

Programs will disappear before anyone is charged. That is the design. The order works through pre-enforcement chilling. By the time a court examines whether the definition is constitutional, (Arnold & Porter has identified real vulnerabilities in its legal authority) the career fair tables will already be gone. The mentoring programs will already be dissolved. The supplier diversity pipelines will already be dismantled. Winning the legal argument in 2028 does not rebuild what was abandoned in 2026.

The question underneath

I started this essay on March 26, 2026, with a compliance officer who could not tell whether recruiting at Howard University was still legal. I came back this week expecting to find that the attorneys had answered her question.

They have not. Every law firm analysis I read, from the largest firms in the country, firms that represent the companies making these decisions, says the same thing in different words: the definition is broader than existing law, the boundaries are unclear, and the enforcement risk is real.

What they have answered is a different question. They have mapped the enforcement machinery. And what that machinery reveals is that the order does not need to answer the compliance officer’s question to work. It needs the question to stay unanswered. An answered question is a line you can stand on. An unanswered question is a space you retreat from.

This is true regardless of what you think about DEI. If you believe diversity programs have gone too far, you should still want the rules to be clear, because vague rules give the government power that outlasts any single policy position. The next administration can use the same machinery for opposite purposes. If you believe diversity programs are essential, you need to understand the mechanism that is dismantling them, because the mechanism is not the headline. The mechanism is a single word in a contract clause that most people will never read.

I teach people how to ask a specific question when the words in the room are carrying more than one meaning: What do you mean?

The compliance officer in Northern Virginia needs to ask that question. Her general counsel needs to ask it. The contracting agencies inserting the clause need to ask it. The DOJ officials reviewing whistleblower claims need to ask it.

Nobody built a place for that question to land.

Thirteen days from now, Friday, April 25, 2026, the clause goes into every federal contract in the country. The word “discrimination” will mean what it has meant for sixty years, and what it has meant for seventeen days, at the same time, in the same sentence, and nobody signing the contract will be able to tell which meaning counts.

The career fair table at Howard University is a real thing that helps real people get real jobs at real companies. Whether it survives the next thirteen days depends entirely on whether the person making the decision can tell what “discrimination” means.

Right now, she cannot. And that is the point.

I wrote the first version of this on Thursday, March 26, 2026, the day Executive Order 14398 was signed. I did not publish it because I wanted to see what the legal professionals would find when they read the fine print. What they found was the word “material,” and what “material” activates is a bounty system most Americans have never heard of. The mechanism matters more than the headline. It always does.

Jerry W. Washington, Ed.D., is the founder of What Time Binds and the creator of the Meaning Repair as Cognitive Infrastructure (MRCI) framework. He is a retired Marine Corps Master Sergeant, a USC Rossier School of Education graduate, and an instructor at UCI Division of Continuing Education.

Start Module 1 of the Meaning Repair for High-Stakes Teams course — completely free, no paywall.

Previous What Do You Mean? entries: What Do You Mean, “Rally Around the Flag”? | Version Control for Democracy | The Inverse Problem