Evidence

What counts as evidence in nasab

Not all genealogical claims are equal. How documentary, physical, testimonial and scholarly evidence differ in weight — and what makes a lineage claim checkable.

A genealogy is a set of claims about who descended from whom. Like any other claim, each one is only as good as what supports it — and the difference between a family tree that holds up and one that does not is rarely the research effort. It is whether the person who built it recorded why they believed each link.

Muslim intellectual tradition is unusually well equipped for this. The science of isnadإسناد — evaluating a report by scrutinising its chain of transmission rather than only its content — was developed precisely to answer the question of how much weight an inherited claim deserves. Genealogy asks the same question about a different kind of chain.

The four kinds of evidence

Documentary

A record created at or near the time of the event it describes, by someone with a reason to get it right. Marriage contracts, land and revenue records, endowment deeds, civil registration, migration paperwork, school and military records.

Documentary evidence is the strongest category, but its weight still depends on proximity. A birth registered the week it happened is a different thing from a birth year written on an identity card issued forty years later on the strength of the holder’s own estimate — and the second is extremely common in records from populations where registration arrived late.

Physical

Gravestones, inscriptions, endowed buildings. A headstone naming the deceased and their father is a durable, contemporaneous, publicly visible statement of a genealogical link, which makes it unusually good evidence.

The caveat is that stones are replaced. A grave marker recut in 1990 for an ancestor who died in 1890 records what the family believed in 1990. Photograph the whole marker, including its condition, and note anything suggesting a later hand.

Testimonial

A named living person stating what they witnessed or were told. This is where most recent family history actually comes from, and it deserves more respect than document-centric genealogy usually gives it — a daughter knows who her father was without needing a certificate to prove it.

Weight depends on proximity and on the number of independent chains. Someone reporting their own parentage is close to unimpeachable. Someone reporting a great-grandfather is transmitting a report they received, and the question becomes who they received it from. Two relatives who heard it from the same grandmother are not two independent sources; they are one source, twice.

Scholarly

Published genealogical compilations, registers kept by lineage bodies, and citations in biographical dictionaries. The Islamic world produced a very large literature of this kind — tabaqatطبقات works, regional nasab compilations, registers maintained by offices responsible for recording descent.

These are valuable and they are not automatically decisive. A compiler working in the seventeenth century about a tenth-century figure was doing the same thing you are, with less access. Treat a published chain as a strong report with an identifiable author, which is precisely what it is — and check whether the compiler names their own source.

What consistently fails

• Unsourced online trees. Aggregation sites copy from each other. Ten sites agreeing usually means one unsourced entry copied nine times, and the repetition is often mistaken for corroboration.
• Name collision. Common names plus a shared region plus an approximate era is not an identification. Two men named Muhammad ibn Abdullah born in the same district within twenty years of each other is an ordinary occurrence, not a coincidence worth building on.
• Aspirational attachment. Connecting a documented local line to a famous distant one across a gap that has no evidence in it. The join is usually the weakest point in the entire tree and almost always the least examined.
• Circular sourcing. A book citing a family record that was itself compiled from the book.

How Shajrah handles this

Every public record carries a confidence score derived from what is attached to it and who has corroborated it, and the record displays that score rather than hiding it. A person with a photographed grave, a marriage contract and three co-signing relatives reads differently from a name someone typed in — and both are allowed to exist, clearly labelled.

Records that predate documentary practice are marked as historical rather than being forced through an evidentiary standard they cannot meet. The prophetic lineage is the clearest case: it is traditional material, and it is labelled as such rather than presented as evidence-backed. You can see how that is handled on the prophetic lineage page.

Disputes are public. If you believe a record is wrong, you raise it against that record and the discussion is visible to anyone looking at it, along with its outcome. A genealogy where disagreements are hidden is a genealogy where the reader cannot tell which claims were ever tested.