* Born out of wedlock

[tatewise]

The full GEDCOM 5.5 description of a Family Record on Page 24 is:

The FAMily record is used to record marriages, common law marriages, and family unions caused by two people becoming the parents of a child. There can be no more than one HUSB/father and one WIFE/mother listed in each FAM_RECORD. If, for example, a man participated in more than one family union, then he would appear in more than one FAM_RECORD. The family record structure assumes that the HUSB/father is male and WIFE/mother is female.
The preferred order of the CHILdren pointers within a FAMily structure is chronological by birth.

So it certainly goes beyond formally Married couples.

Some definitions for 'family' say it is a group of individuals who share a legal or genetic bond.
So, even a 'one-nighter' resulting in child would share at least a genetic bond and thus classed as a 'loose' family.
To build a biological/genetic 'family' tree, the parents and the child are a type of family.

[ColeValleyGirl]

So No child == No family. Which is of course nonsens if you're doing Family History rather than a narrow lineage-based genealogy.

[tatewise]

Sorry Helen, but where did anybody say No child == No family? I agree, that would be nonsense.

[davidf]

If you record that a couple got engaged, you create "a family"!

[ColeValleyGirl]

The FAMily record is used to record marriages, common law marriages, and family unions caused by two people becoming the parents of a child

Depends on whether they use Oxford commas or not. (And of course nowhere in the spec is that made clear.)

Is this "marriages common law marriages and family unions" all "caused by two people becoming the parents of a child" (so a childless relationship is not a family); or "marriages common law marriages" and "family unions caused by two people becoming the parents of a child".

A common sense reading suggests the second but having spent a portion of my career dealing with both International Standards for various things technical and writing unambiguous (safety related) software specifications using Z notation, I have never been impressed with the clarity of the Gedcom specs and both readings are grammatically valid in the absence of guidance.

But the definition (however you read it) excludes English childless unmarried couples from being families, and they satisfy none of the criteria (common law marriage not existing in the country.) And by a strict reading of the spec engagement =/= family.

[tatewise]

We all know that the GEDCOM specification is riddled with contradictions, inconsistencies, and only partial explanations.
It is those that GEDCOM 5.5.5! (17121) is trying to rationalise.
I mentioned the GEDCOM description of Family records to show it does not require the couple to be formally married.
Clearly the description is not exclusive, and does allow unmarried childless couples, because Engagement and even Marriage Banns/Contract/Licence are standard events applicable before marriage and potentially childless, with no guarantee that a Marriage will follow.

[AdrianBruce]

... if Simon and Susan have a "on-nighter" and go their separate ways but Stephen is the result (acknowledged by both), how should these three be entered - given there is no "two parent" family and it would be untrue to say that it is a "one parent" family? ...

Why is there not a two-parent family? There might not be a social family but if the GEDCOM concept of family also hangs on and around the biological aspects, then a two parent GEDCOM family is exactly what there is. More usefully, but only if the reader is IT-literate, the GEDCOM family includes the concept of a biologically based triplet - three entities whose relationships are biological. Mother/Child/Father is the triplet.

Of course, the GEDCOM family also includes the ability to record spouse/spouse doublets and squashes them into the triplet as Mother(spouse)/Child/Father(spouse), and then mashes into the mix extra children and .... All a bit messy if you like purist data modelling (I do) but actually fairly workable until you start thinking about it in IT terms.

If you want to see the mess that thinking about it too deeply gives you, you should spend some time with FamilySearch FamilyTree where they've (commendably?) attempted to separate the biological triplet from the spouse/spouse doublet and ended up - since the "parents" appear in both - with a structure that does allow the separation but is seldom used in that manner because it's virtually impossible to explain in text or diagrammatically.

NB - I dislike referring to the biological triplet of Susan/Stephen/Simon as a Family - but in GEDCOM terms, that's what it is, and until someone comes up with a better term and / or better data model, that's what we're stuck with.