Identifying biological parents

[tychonievich]

I had forwarded to me a longer email that contained the following ideas, which I've re-expressed to remove the ideas from the context of that email.

---

We use 5.5.1 FAM records to express biological relationships: HUSB for biological father, WIFE for biological mother, CHIL for biological child. We find some applications don't enforce the biological gender of HUSB and WIFE and fix those files on upload when possible, produce errors if not. We also use extensions _FREL and _MREL to note the type of relationship, treating some values as indicating biological relationships.

Version 7.0 explicitly allows HUSB and WIFE to be used in other ways; for example, it says

Sex, gender, titles, and roles of partners should not be inferred based on the partner that the HUSB or WIFE structure points to.

Because of this, with 7.0 the FAM records no longer can WIFE and HUSB records be used to determine biological father and mother. How do we determine biological parents with GEDCOM 7.x?

[tychonievich]

The changes in the text describing FAM in 7.0 were an attempt to document common practice with 5.5.1 rather than to provide new meanings. We identified some applications that explicitly used what we documented already and many more users of applications who imposed this practice on applications that did not control how it was used.

Biological sex as identified at birth can be found with the SEX structure.

7.0 keeps two standard structured first introduced in version 5.5 that can represent biological birth:

INDI.BIRT.FAMC is identified in the 7.0 specification as indicating the family of biological birth: in particular it says "biological parents can be shown by a FAMC pointer subordinate to the birth event." Note that it does not define that field as only meaning that; perhaps we should add that?

INDI.FAMC.PEDI can indicate the type of relationship the child has to the family, with one value being BIRTH. I believe the people who added this to 5.5 thought it meant biological, but we know some applications have interpreted in other ways for surrogacy and prearranged adoptions. If we are not sufficiently happy with the BIRT.FAMC option, we could add an unambiguous GENETIC option to the PEDI enumset in 7.1.

I thought we had text in the spec that recommended that HUSB and WIFE be used for biological father and mother, respectively, when the FAM has that structure, but I can't seem to find it. If it indeed is not there, perhaps we should add that recommendation as a patch?

[albertemmerich]

Using _MREL and _FREL in FAM records is against rules of spec. The relation in between parents and children are documented in child's INDI record as shown by Luther. Extended tags should only be used if the spec does not allow to export the data. We have applications in the wild which use _MREL and _FREL and at import modify adopted children to "natural" children. On reimport to applications following the spec the information is wrong (if the reading application tries to read _MREL or _FREL) or missing (when ignoring these bad extensions).

[albertemmerich]

The 7.0 spec has this note: "It is known that some users have interpreted BIRTH to mean “genetic parent” and others to mean “social parent at time of birth”. Definitions differ in many circumstances (infidelity, surrogacy, sperm donation, and so on). Hence, applications should refrain from asserting it has either meaning in imported data." PEDI BIRTH is not clear. In most cases (using church books or civil registers) it is used with the social father. Spec calls it "Family structure at time of birth" Perhaps we should at an ENUM value for PEDI: GENETIC, to be used, if the "biological" father is confirmed by DNA test. See Luther's comment.

[tychonievich]

On @albertemmerich 's point that _MREL and _FREL are not to be used, this is an application of the spec's Extension versus Standard policy that if a standard tag provides the information, it shall be used in lieu of or in addition to an extension tag providing the same (or a superset of) that information. Thus if _FREL or _MREL have information that cannot be provided by a standard tag (I don't know that it does) it can be included, but only along with the corresponding FAMC, PEDI, and so on.

[tychonievich]

Discussed 2023-08-10

• We should probably address single parent-child relationships (such as _FREL, _MREL, GEDCOM-X, etc) vs couple-child relationships (such as FAMC) in the Technical FAQ
• At least 3 types of birth families exist: genetic (e.g. as checked by DNA), accepted (e.g. as on birth certificate), and social family structure (e.g. who is caring for the child). Neither 5.5.1 nor 7.0 have ways of programmatically distinguishing these. In 7.1 we should add PEDI enumeration values for these (GENETIC, LEGAL, and the current BIRTH/ADOP/etc)
• We should probably add test files for these various cases

This is a complicated topic; if we have failed to consider some case, please let us know by follow-up

[jl5000]

With all of this discussion, it has caused me to question my existing understanding of INDI.FAMC.

If a child was born of a biological mother and father that never lived together or behaved as a family, would you create a FAM record containing the 3 of them, or would you create two FAM records, one for each parent, adding the child to both of them?

There seem to be a variety of ways of representing situations like this that it probably does need more explanation.