Analysis

The record-state residual: three distinguishable states of the record holding about 44 percent between them

Eliminating named households is not the same as clearing the record. The same-name sweeps found more than forty distinct John Gurney heads of household in England between 1600 and 1670, and English parish-register coverage for the period is patchy enough that others existed unrecorded. The 1662-1666 London hearth-tax cluster makes the point concrete: eleven Gurney households in London alone, none of them surfaced by any candidate search, because a search built around a name, a wife and a child set cannot see a household the registers never entered. The residual is not one undifferentiated unknown but three distinguishable states of the record, and they call for completely different work: no record survives or is indexed anywhere, at about 22 percent; documented but never assembled into a person — scattered single events never linked, or records wrongly merged into another household — at about 17 percent; and documented but sitting in a source or class not yet searched, at about 5 percent. Together those three hold about 44 percent, more than every named candidate combined, and they are the principal reason no named candidate rises above about 11 percent.

Confidence: Moderate-high

Topic file location: Identity — assessment