reading the marks
How good is a link?
Every parent on a person's page carries a small mark saying how well that particular parent-to-child link is proved. This page says what each mark means. It is reference, not news — what's new records when things changed. Figures below are counted from the published tree of 6 September 2026.
Why a family tree needs this at all
A tree drawn as a diagram tells you nothing about how much of it to believe. Every line looks equally confident, and the weakest link in a chain of ten generations is invisible. Genealogy has a standard for this — a link is proved when a reasonable search has been made, the sources are cited, the evidence is analysed, conflicting evidence is resolved, and the conclusion is written down. What follows is that standard, said in plain words, and applied link by link rather than to the tree as a whole.
92 of 220 marked parent-to-child links in the published file have been read under this vocabulary so far. The rest carry the older plain check mark, and the review is working outward from the generations nearest the present. A parent with no mark at all has not been assessed yet — which is a different thing from having been assessed and found wanting.
The marks
Proven on a record
A document read for this project — the register page, the certificate, the will itself, not an index or a summary of one — states the relationship outright. The strongest thing this tree can say about a parent.
68 parents on the published tree
Proven by argument
No single record says it. Two or more independent records, set against each other in writing with every conflict dealt with, leave no room for another answer. The reasoning is written down before the link is marked, and the note on the person's page carries it.
24 parents on the published tree
Proven — graded before this vocabulary existed
Marked proven in the working database, on evidence, before the six-class vocabulary was adopted in September 2026 — and not yet re-read under it. It is not a weaker claim; it is an ungraded one. The review is working outward from the closest generations.
102 parents on the published tree
Asserted, and under review
This one is the honest awkward case. The link is asserted in the tree, the question has been looked at, and what is in hand reaches neither of the two bars above. The note on the person's page names the cheapest document that would settle it. Until that document is read, the mark here says so rather than saying proven.
15 parents on the published tree
Not proven — one informant
A single uncorroborated statement by somebody who need not have been present — a death certificate's informant line, a compiled family sketch. Looked at, and declined. The link stays in the tree because it is probably right; it is not claimed as proven.
3 parents on the published tree
Refuted by a record
A document read for this project contradicts the link.
none on the published tree
Disputed
Records disagree with each other and the disagreement is not yet resolved.
6 parents on the published tree
Contradicted
The link is in the tree — usually inherited from an older published pedigree — and the evidence read here runs against it. The case files tell these stories.
2 parents on the published tree
Where the reasoning is
Under Research notes on a person's page, kept verbatim: what was read, what it proved, and what stays open. Where a link is proven by argument, that note is the argument — including the part naming what would overturn it. The case files tell the longer stories, and the tree as a file is downloadable in full.
One caveat about the file: a GEDCOM carries conclusions, not arguments. If you import this tree into your own, the marks on this page do not travel with it.
If you think a link is wrong
Say so. A tree published with its weak links marked is inviting exactly that. The main site has a way to get in touch, and a correction with a record behind it is the most useful thing anyone can send.