The Power to Always Return to the Source: Tying Every Claim to Approved Evidence

"I'm pretty sure it said so somewhere" is the most dangerous sentence in a material. The power to return to the source means being able to walk every claim you wrote back to its approved original evidence. Not memory, not hearsay, but opening the original page.

1. Returning to the source means "finding your own way back"

Think of a mountain trail. On the way up, you reach the summit by following signs and companions. But whether you truly know the path shows on the way down, alone. Can you explain which fork you turned at? A claim in a material is the same. When you write "this drug works on X," can you walk that sentence back, on your own feet, to the approved source data (clinical results, the package insert, and other officially recognized evidence)? That is the power to return to the source.

Approved evidence here means "the original information that has been officially verified and cleared for use." Someone quoted it in a document, a manager said so, the previous version had it written that way: these are not evidence. They are shadows of evidence. No amount of shadows equals one page of the original.

A claim is "grounded" only when its author can find the way back alone. A claim that cannot return, however correct it looks, floats in the air.

In the two axes from Part 2 (fidelity to fact x design for delivery), this power sits at the center of fidelity, that is, grounding. However skillful the delivery, a claim that cannot return to its source data has lost its foundation. Grounding is the ceiling that caps design.

2. The greatest danger is "secondhand citation," a game of telephone

Recall the children's game of telephone. The first child says "big red dog," but after five people it has become "big blue cat." Each person meant no harm; each just rephrased a little. The same thing happens in material-making through secondhand citation: copying numbers or wording from a secondary document (another text that summarizes or quotes the original) without seeing the original itself.

Why is this dangerous? The person who made the secondary document may have already rounded things off. "About 70%" may have been "68.2%" in the original, with the condition "only in a specific patient group." Copy it with that condition dropped, and it quietly mutates into a different claim. The table below shows how the gap opens along the way.

By the third row it is a different claim. You may not call it false, yet the impression the reader receives has shifted far. So the power to return to the source does not stop at "someone's summary." It always descends to the single page of the original. It does not get off at an intermediate station.

3. "Embellishment" is not a lie, but it drifts slightly from fact

The plating of a dish makes this clear. With the same ingredients, add lighting, a wisp of steam, change the plate, and it looks far more appetizing. The taste has not changed. Yet in looking "better than reality," the expectation it creates has shifted. Embellishment in a material is writing no lie, but, through word choice, making the fact look stronger or better than it is.

If the original says "a reduction in symptoms was observed" but the material says "symptoms dramatically disappeared," that is embellishment. "Was observed" and "dramatically disappeared" do not point to the same fact. The power to return to the source checks each word against the original, asking "is this strength one the evidence permits?" Adjectives and adverbs ("dramatically," "greatly," "immediately") are the doorways through which embellishment slips in.

"As long as it is not a lie" is not the standard. The moment you exceed the strength the original states, you begin to drift from fact. Embellishment is "skillful deviation," and the more skillful, the harder to notice.

Recall the danger zone of the concept model. The most dangerous was "high design x low fidelity = persuasive misunderstanding." Embellishment is exactly this. The better someone writes, the more they unconsciously dress things up. So the skilled writer above all needs the habit of pulling their own words back to the original.

4. The stages: from fixing on request to building a system that returns

Think of driving. At first the instructor says "brake now" and you press. Next you read the signs and stop on your own. Then you grasp from the road's structure why slowing is needed. Finally you can design a road on which accidents rarely happen. The power to return to the source grows through the same four stages (L1-L4).

The L4 "ledger" is nothing complicated. It is a list with each claim of the material on one line, and beside it "which approved data, and where in it." With this, anyone can later retrace the path. Grounding is guaranteed by the system, not by individual memory. That is the L4 mindset.

5. Why grounding is a "floor," not an "ornament"

Think of a house's foundation work. The foundation is invisible in the finished house. Guests praise the wallpaper and lighting, never the foundation. Yet if the foundation tilts, the loveliest interior tilts with it. The power to return to the source is this foundation. Hard to see from the front, it holds up everything.

In the logic of pass or fail, grounding is a "non-compensable gate." Non-compensable means no other strength can make up for it. Persuasive appeal, beautiful design, clear explanation: each has value, but the single defect of being unable to return to the source cannot be patched by any of them. An error painted over with persuasion only causes greater harm. So grounding is treated as the "floor" (a necessary condition) that must be met before any contest of scores.

A floor's condition is that you do not fall through it; it is not a place to be praised. Being able to return to the source is not excellence but the ordinary minimum. The one who never lets go of that ordinary thing builds trust.

Note that this differs in role from the source-checking of the reviewing side (the reader's check). Review confirms a finished material from the outside. The creator, before release, returns their own claims to the original from the inside. Even looking at the same original, they stand in opposite positions. When the creator firms up this point, review becomes that much faster and surer.

The power to return to the source is not a flashy one. Readers usually praise clarity or looks; rarely do they notice "this material can properly return to its original." Yet the moment this power is missing, every other strength tilts along with the foundation. Do not stop at memory or hearsay; catch secondhand citation; pull embellishment back one word at a time. And make it a system that links each claim to its source one to one, not a matter of personal memory.

Part 4 turns to the power to design balance: handling benefit and harm with equal weight. Once the ability to return to the source is firmed up as a floor, how do we lay out that evidence fairly? After grounding comes balance.