01 — Sourcing

Where the Facts Come From

The verified facts in The Record are drawn from wire services (AP, Reuters), official government records, court filings, legislative texts, and primary documentation. When a government agency releases a number, that number is sourced to the agency. When a court issues a ruling, the ruling is cited. When outlets conflict on a basic fact, the conflict is disclosed and the primary source is identified.

No unnamed sources appear in The Record. No hedged attribution. No "sources say." If a claim cannot be traced to a named, verifiable source, it does not appear.

Sources are listed at the bottom of every story in monospace text. The reader should be able to verify every claim independently.

02 — Story Selection

What Gets Covered and Why

A story earns the full treatment only when right-leaning and left-leaning outlets tell meaningfully different stories about the same event. "Meaningfully different" means different facts emphasized, key context omitted, loaded language deployed, or fundamentally different narratives constructed from the same raw material.

If both sides covered an event the same way, it either goes to Consensus Watch (if the agreement itself is surprising) or it gets skipped. The publication does not force stories into its format. It does not manufacture conflict to justify its own structure.

The monitoring span covers print and digital outlets across the spectrum: Fox News, New York Post, Daily Wire, Washington Times, Wall Street Journal editorial page, and Breitbart on the right. CNN, MSNBC, NPR, New York Times, and Washington Post on the left. Broadcast commentary is tracked separately from written reporting because the same outlet often tells two different stories depending on the medium.

03 — Framing Analysis

How Media Coverage Is Examined

Right Said and Left Said are not summaries of coverage. They are specific, attributable observations. Each section must contain at least one named outlet doing a specific, verifiable thing: a headline it ran, a quote from a host, a word it chose, a fact it buried, a framing decision it made.

The reader should be able to check each example. If Fox News is cited for a headline, that headline existed. If MSNBC's host is quoted, that quote is from the broadcast. Framing analysis is never assumed, never generalized, never fabricated. If specific coverage cannot be found and verified, it is not included.

Print coverage and broadcast commentary are distinguished. A newspaper's measured reporting and the same outlet's primetime show may tell contradictory stories. When that gap exists, the publication notes it.

04 — The Gap

Accountability for Both Sides

The Gap calls out the right's distortions first, then the left's. This is a deliberate structural choice. Right-leaning readers who perceive bias will disengage early. Presenting right-leaning accountability first signals that the publication takes their media ecosystem seriously before asking them to read criticism of the left.

Every claim in The Gap is documented against the factual record. "Both sides spun it" is not an acceptable statement. The specific distortion is named. The documented fact that contradicts it is cited.

The publication follows evidence, not a quota. Some weeks one side distorts more than the other. The Gap reflects what the evidence shows. Even-handed does not mean both-sides-equal.

05 — Claim of the Week

Sunday Editions

The Claim of the Week identifies the single public statement from the past seven days with the largest gap between what was said and what the documented record shows. Any politician. Any party. Any public figure.

Selection favors claims checkable against official statistics, court records, budget documents, or primary documentation. The gap must be factual, not philosophical. A policy disagreement is not a false claim. A number that contradicts a government's own data is.

Claims are never labeled lies. The publication presents the claim, the documented record, and the distance between them. The reader decides what to call it.

06 — Democracy & Governance Watch

Sunday Editions

This section monitors ten categories of institutional health: erosion of independent oversight, politicization of law enforcement or intelligence, actions targeting press freedom, use of institutional gridlock as governance strategy, actions affecting electoral integrity, unusual concentration of executive authority, use of government agencies for partisan purposes, suppression of lawful dissent, actions affecting judicial independence, and normalization of political violence.

All branches are monitored. All parties. All levels of government. Every entry requires documented evidence from the current week's events. If the tracker consistently flags actions from one political direction over consecutive weeks, a transparency note acknowledges this and states whether the pattern reflects documented asymmetric behavior or a potential limitation of the framework itself.

07 — Consensus Watch

When Agreement Is the Story

Consensus Watch appears when left and right outlets converge on a story in a way that is genuinely surprising. Both sides calling for the same investigation. Bipartisan outrage over the same event. Both sides acknowledging facts that contradict their usual narratives.

Routine agreement is not included. Consensus Watch exists specifically to highlight the moments when the partisan framing machinery has nothing to grab and the media briefly agrees on what happened.

08 — Developing Stories

When Facts Are Not Settled

When a major story is still unfolding and facts remain unsettled, it receives the Developing tag. The Record becomes "The Record So Far." Only confirmed facts as of publication time are included. If framing divergence has not yet solidified, the publication says so rather than forcing premature analysis.

Developing stories are tracked across editions until resolved. The tag is removed when facts settle and the full four-section treatment can be applied with confidence. Stories are never tagged as developing to manufacture continuity across editions.

09 — Verification

The Adversarial Review Process

Before publication, every edition undergoes an independent adversarial review that challenges each factual claim through separate research. This process does not trust the production draft. It re-verifies. When it finds errors, it produces a correction report that is implemented in full and logged in the edition record. The publisher reviews the corrected edition as the final check.

The full production and verification system is described on the AI and Production System page.

10 — Corrections

When We Get It Wrong

Errors are corrected and logged in the edition's footer with the date, what changed, and the corrected value. If a correction changes the substance of a story's framing analysis, the edition is republished with the correction noted prominently. The archive reflects the corrected version, not the original error.

To report an error: Email Brian.