These research notes are provided as-is and contain supplementary working research.

This page expands an Evidence link in Identity — the shape, the matrix, and the two families: the findings, sources, and qualifications behind the marked passage.

Analysis

Eight corrections to the shape, and a ninth rule from the elimination side

The shape as first specified needed eight corrections, in descending order of how much each changes a score. Margaret belongs on the mother axis as a substantial soft hit. The Mary attribute has a primary bound and the Richard and John attributes do not - Mary Gurney married Daniel Shed at Braintree in 1647, so at a minimum marriage age of sixteen she was born by 1631, while Richard about 1630 and John about 1633 rest entirely on compiled estimates with no primary record behind them; this matters directly, because Hitcham's Mary of January 1630/1 satisfies the primary bound exactly while looking three years off the traditional date. The two marriage attributes are mutually exclusive, so scoring both double-counts one marriage. The paternal-grandfather attribute is a derived prediction rather than an independent observation, conditional on Richard being the eldest son, so a hit on it is a hit on a hypothesis and a miss is weak evidence. A departure attribute should be added - the household's records stop between 1635 and 1641 with no burial or probate for the head - but only with its coverage control attached, since two false departure gaps have already been produced by omitting it, and on a properly controlled test only Ackworth currently passes. A penalty is needed for children with no colonial counterpart, since as specified the shape can only reward matches and never penalise surplus. The surname attribute is a weighted scale rather than a single tick. And Peter, Isaac and Sarah must be excluded from scoring rather than treated as missing children, because their absence from English registers is expected. A ninth rule arrived from the elimination side: three undated role-row households, one of them earlier described as the single most important loose end in the work, were dated on a second instrument and resolved to 1687, 1704 and 1710 - so an undated role row is not evidence of an in-window household until it is dated elsewhere.

Confidence: High

Why this conclusion

Qualifications and conflicts

Sources

Appears in

  • John Gurney identity case file — Section 1's note following the fact table carries two of the corrections: Peter and Isaac as American-born and Sarah as doubtful on her own account.

Topic file location: Identity — the shape, the matrix, and the two families

Also covered here

Technical details
  • Evidence marker: G13-PM-000129
  • Primary research item: G13-RI-000227
  • Graph revision: 381
  • Machine-readable JSON: /assets/g13-graph/marker-bundles/G13-PM-000129.json