Preserve what arrived
Keep each received trace—the record as it actually arrived—before interpretation changes it.
Company method
Test the bridge before you trust the claim.
LFR examines how records come to support claims about an object, process, or event—physical or institutional. The first question is not only whether the number is correct, but whether the record still refers to what we think it does.
A record may be authentic and a calculation reproducible while the represented relationship remains unestablished.
Boundary. This page describes the company method. Any field-specific claim still has to earn its authority from its own evidence, contracts, and release checks.
Four actions
What LFR does.
Four actions keep a claim connected to the records that are meant to support it.
Test the bridge
Check identity, time, and meaning. A declared name is not verified identity.
Declare the transformation
State what was done to the record and which assumptions entered.
Qualify the claim
Say only what the evidence supports, and keep unknown or refused claims visible.
From event to claim
One event can leave more than one record.
The records may agree, conflict, or leave part of the event unseen. LFR preserves what arrived before deciding what any record can support.
Records left by the event
TEST THE BRIDGE
Does this record still refer to the thing we think it does?
We test whether the record can be linked to the claimed object, process, period, or institutional relation.
- same object?
- same period?
- same meaning?
- same units and sign?
- missing parts accounted for?
source exists ≠ object identity establishedtimestamp exists ≠ time relation establishedcalculation reproducible ≠ represented relationship established
The transformation may now proceed with its inputs and assumptions declared.
The record may be real, but its identity, timing, or meaning is still uncertain.
The claim path stops. This does not mean the claim is false.
Why stopping matters
Stopping is part of the method.
If identity, timing, or meaning cannot be linked, LFR leaves the relation unresolved or refuses the claim. It does not repair the gap with prose.
A missing record is not a null result. A null result says only that the stated test did not establish the tested change or relationship under its declared conditions and observation limits. It does not establish that nothing happened.
Instantiated fields
The method is shared; evidence is local.
GB Grid and Battery/BESS are field-specific implementations of this method. They do not inherit authority from this page; they earn it from their own records.