ProofStatement has no meaning in Attribution

[EssyGreen]

Attribution is effectively documenting the creator of the Resource. (Which is also done via the Dublin Core description so this seems to me to clash.)

Proof, by definition, is the justification for some sort of statement, usually something like why A = B. But, whilst A could be assumed to be the Resource that the attribution is contained in, there is nowhere to define B so it doesn't make any sense.

In a genealogical sense the Proof is a weighing up of the evidence to form a rational conclusion. To use FamilySearch's own definition (see https://www.familysearch.org/learn/wiki/en/Genealogical_Proof_Standard):

There are five elements to the Genealogical Proof Standard:

• A reasonably exhaustive search has been conducted.
• Each statement of fact has a complete and accurate source citation.
• The evidence is reliable, and has been skillfully correlated and interpreted.
• Any contradictory evidence has been resolved.
• The conclusion has been soundly reasoned.

Thus there must be a collection of relevant sources representing evidence/citations for the ProofStatement to make any sense. These are currently held in the "Conclusion" object. Ergo the Confidence and ProofStatement should be properties of the Conclusion and not of the Attribution.

[EssyGreen]

Another way of resolving this could be to have an iProvable interface instead of a Conclusion object. That way anything that exhibits the following properties can be proven

• collection of sources
• ProofStatement
• Confidence

This would enable greater flexibility since it would separate the inheritance (what kind of object this is, what sort of shape it has) from its function (what things can this object do/interact with).

[stoicflame]

Thus there must be a collection of relevant sources representing evidence/citations for the ProofStatement to make any sense. These are currently held in the "Conclusion" object. Ergo the Confidence and ProofStatement should be properties of the Conclusion and not of the Attribution.

Sorry, didn't catch the reasoning there. Why is a problem that the proof statement is associated with the person who made it? Seems to make sense to me....

[EssyGreen]

I don't have a problem with the Proof being associated with the person who made it but that is implicit by the Attribution in any resource. It's the evidence which is missing and the other half of the equation. For example, I (contributor) could prove that England (place) is in the UK (place), but I can't "prove England". And neither holds any water without the references to the evidence to support my argument. Since none of these are available in the Attribution then the proof is meaningless in the context of the Attribution.

[stoicflame]

Since none of these are available in the Attribution then the proof is meaningless in the context of the Attribution.

But the proof statement isn't outside the context of the conclusion. Attribution has no meaning outside the context of the data being attributed.

Maybe I can ask from a different angle: why is it okay to have a contributor reference (or timestamp, or whatever else is there) on attribution if it's not okay to have a proof statement on it?

[EssyGreen]

Let's take some examples of things which are Resources and hence have Attributions:

• a Person/Record/Field/Relationship/Collection (inherited from the GenealogicalEntity) - the contributor is the researcher say - no problem; the proof is based on the collection of Sources in the base GenealogicalEntity so again no problem.
• a Field (e.g. AgePart, DatePart, Fact) in the Record Model - the Contributor is the interpreter who transcribed/translated/interpreted this particular field; there is no evidence (only the source from which the data was taken) so the proof can only be something like "It looked like this in the Source". This may be deductive reasoning (or in my terminology a Hypothesis) but it's not what most genealogists would consider to be proof since there is only a single source to base it on.
• a Note - the Contributor is the researcher say who wrote the text in the Note but there are no sources to be used as evidence so the proof can only equate to "because I said so"
• a Relationship (presumably in the Conclusion Model but no longer inheriting from a Conclusion for some reason) - the Contributor is the researcher who documented the Relationship; there are no sources to support the Relationship so the Proof can only mean "because I say so"

If I'm missing something please illuminate :)

[stoicflame]

(Just doing some triage.)

The notion of proof statement was refactored and clarified at #144 et. al.