Onomastics: where to store name etymology?

[pgerlier]

While collecting genealogical data, I am also interested in onomastics, i.e. the etymology of proper names.

As is frequently the case, an old name may be victim of corruption, usually because of the fundamentally oral society in which people lived, their pronunciation and the clerk not being accustomed to the name repertoire.

I heavily use the “Group as” field to gather all variants under a common conventional name. However, this does not allow me to comment and explain the origin of names (from a linguistic and historical point of view) because there is no unique “central” record for a name.

I am presently exploring an alternative way of storing Gramps data. I am considering creating a new TABLE (Gramps object) to replace the “Group as” field by a full record on its own so that this record can have notes, attributes, tags, privacy, …

Would this be a valuable extension? Are others interested?

[PLegoux]

Such object would be interesting, particularly when two groups for the same kind of name are dissociated because patronyms are not related and there is a need to explain it

[emyoulation]

First, any kind of a surname etymology modeling is going have a huge cultural bias. It seems like each Origin Attribute would have different requirements for how Annotations will pertain to another entry.

For some purposes between Individuals, the Associations is useful. I regularly create Eponym/Namesake entries for Associations. That help during direct inspections but there are no analysis tools since these are not built-in Association types. (Although @GaryGriffin was kind enough to create Sync Associations tool to simplify keeping Reciprocal Associations from being lost.)

BTW, on the webpages found with a quick Googling, there were Onomastic branches about place names (toponyms) and personal names (anthroponyms). I did not see a branch concerned with Surnames.

Perhaps some groundwork would be informative before extending the Gramps Data Model?

It is hard to see the progression of variants with our current tools. They are too broad. The Gramplets do not allow only Ancestor of the Active Person (or some other Filter) to be considered. So there’s too much noise obscuring any pattern.

The Groups in names has only rudimentary sub-sorting. Several people have noted that it would be more useful if the sub-sorting within a Group listing ignored the top-level. (i.e., if all the "Edwin"s within the artificial group of Smith, Smythe & Smyth were listed adjacently rather than separated by surname spelling-variant cliques.)

Names in Gramps

If you have more than a few thousand people in the Tree, the Surname Cloud becomes useless.

Likewise, the Top Surnames is overwhelmed if you store Persons for exclusionary purposes. (i.e., one might collect all the “Smith/Smythe/Smyth” Persons found in a neighborhood, even though expecting only a few to be connected genetically to the central Tree.)

[PLegoux]

@emyoulation

BTW, on the webpages found with a quick Googling, there were Onomastic branches about place names (toponyms) and personal names (anthroponyms). I did not see a branch concerned with Surnames.

In french, anthroponymie, is about study of surnames (may be personal name but mainly surnames), their definition, origin, meanings…

[pgerlier]

@emyoulation

etymology modeling is going have a huge cultural bias

Yes, that why I imagine a “neutral” model. I don’t favour any approach. I want to add Notes and the real job will be in the notes. This can already be done for Places: there is a “central” record (the place) and you can add notes to explain the name (and alternate names). I’d like to have the same possibility, which requires a “central” record for an abstract name (grouping all variants)

@emyoulation

one might collect all the “Smith/Smythe/Smyth” Persons found in a neighborhood, even though expecting only a few to be connected genetically to the central Tree

As an example of this ambiguity, I’ll take Cohen and Berger.

Cohen can be either an Hebraic name (a priest) or a Breton name (oak-tree). Therefore you need two Group as keys Cohen (H) and Cohen (br).

Similarly, Berger may come from Latin (shepherd) or from German (mountaineer), requiring again two Group as keys Berger (fr) and Berger (de).

The group record can be used for any purpose: linguistics as shown, geographic to exhibit independent origins, …

@emyoulation

The Groups in names has only rudimentary sub-sorting

This is another aspect which is not basically related to Group as since it can be used (and is used) even without Group as. And this is a very cultural-sensitive point (see the recent question about Ethiopian names.

I am only in the preliminary steps so any suggestion is welcome.

[emyoulation]

One of the thing that seems ugly is that there should be an interaction between Origin & inheritance.

The current Note in Names seems to be at the Individual level for exception records in Alternate Names.

This would add a Grouping Note? And possibly, an Attribute for future structured flexibility? (i.e., a Lang attribute for those french fr & german de Bergers?)

[pgerlier]

I don’t know for a language attribute. Things are not that simple. In France you meet Berger’s but to pronounce correctly the name \bɛʁ.ʒe\ (fr) or \ˈbɛʁɡɐ\ (de) you need to know the family origin, though most people use the French sound and this is erroneous half the time. Since both names are frequent in one country, would the language attribute be relevant?

@emyoulation

there should be an interaction between Origin & inheritance

I don’t get the point. Could you elaborate a bit?

[emyoulation]

Language is a significant contributor to names. Anglicization is a familiar case I’m alternate name for the American trees. Whether it is a German Zimmerman becoming a Carpenter or a Johann becoming John, a simple 2-letter attribute describes a language shift.

[Mattkmmr]

@pgerlier

I am presently exploring an alternative way of storing Gramps data. I am considering creating a new TABLE (Gramps object) to replace the “Group as” field by a full record on its own so that this record can have notes, attributes, tags, privacy, …

Would this be a valuable extension?

Yes I think it would be great. At the moment I’m storing etymology data as notes, but do not have a way to reference them to surnames or their “group as” field.

[Davesellers]

Forgive my ignorance but could you give me an example of this.

[pgerlier]

What do you mean? Onomastics or my new design for “Group as”?

In the case of onomastics, take my own name Gerlier. It has two possible origins:

• a Provence trade surname
  A “gerl-ier” with sufix “-ier” designating a craftsman is a maker of “gerle” or “jarle”. A “gerle” is a special tub made of wooden staves, two of which (opposite) are longer than the others and pierced with one hole each. Two men can carry the “gerle” with a shaft going through the holes.
  
  
  This special tub is used in the vineyard where the grapes are already pressed before being carried to the cellar.
  This rare name can be found originally around Avignon (Papacy possession in the Middle Ages), but its diffusion to some of its present areas can’t be explained based on trade related to vine work or wine making. In my case, I suspect a forced-enrolled soldier left over (wounded, ill, desertion, …) after some campaign XV-XVI century in Holy Roman Empire founded a branch in a village where he settled totally by chance.
• a Frankish first name
  Here the Germanic form is Gerlach which have been transliterated in old French as Gerlier. The Frankish roots are “gari” for spear and “liut” (a corrupted form being “lach”) for people. Therefore, the Gerlach were warriors. Notable people are count Gerlach I von Nassau (French: Gerlier de Nassau) and his third son Gerlach von Nassau (1322-1371), archbishop in Mainz (Germany).
  Today, there are relatively many Gerlier around Liège in Belgium

Onomastics is the science of roots in person name.

It is then obvious that the same origin can be traced to many variants of the same name. Also, the same written form may refer to disjoint origins as in the example above and should be separated in Gramps. Consequently, it would be more user-friendly to attach the note to some “common factor” of all these names. I thought of the “Group as” attribute which looked like a convenient point.

[emyoulation]

How do you link the eponymous Person for a namesake?

For the Surnames, we have the Origin attribute. But how would I link the eponym(s) for a Given name that is not a Junior/“The Younger”? Particularly, for the custom of honoring grandparents.

As an example, my brother was born on my paternal grandfather and was (reluctantly) named in his honor. But, in the interest of family harmony, his middle name honors the maternal grandfather. Of course, an Eponymical Origin is complicated since Gramps only has a single field for compound Given names.

The Saint/Spiritual name as a cultural standard is another Eponym Origin conundrum. Particularly because the Saint is typically placed before the customary Call name.

Besides Religious/Saint name cultural custom, other non-genetic namesake etymologies are interesting to note. Such as making a public declaration of faith (or trying to inspire the child to emulate a paragon) by choosing a Biblical, Apostal, Heroic/Historical, Political, or Popular eponym.

There are surges of names related to current event. The name is a passing fad which echoes down the generations as the Given name transforms into honoring genetic ancestors instead of a (no longer revered, forgotten or even reviled) historical figure. As a USA example, the Civil War inspired the 1860s Northern/Unionist fad of Col. Elmer Ephraim Ellsworth namesakes. (He was the first conspicuous casualty on the Union side of the War.) There were fading echoes of the name in the 1880s, 1900s & 1920s as generations of namesakes spawned their firstborn males about 20 years apart.

In my own family, a paternal great-grandfather was named after an eccentric itinerant evangelist as a public statement of faith. The boy rebelled against his parents religious evangelism and went by his initials in adulthood.

These etymologies are interesting to track. Moreover, the morphology through inheritance is also complex. If I note that a person was a namesake to honor the immigrant ancestor (rather than his grandfather who honoring the same eponym), it would be helpful if documenting that the immigrant ancestor was inherited. So the immigrant’s eponym would somehow be referenced in both descendant’s stories about their name.

My current approach is to set an “Association” between 2 people using the asymmetrical reciprocals: eponym; namesake
I typically will set the Association in the namesake’s profile. I’ve related the two terms in the SyncAssociations add-on tool and periodically use that to synchronize unmatched reciprocal Associations.

[emyoulation]

Here’s a similar thread (Fame is fleeting. Who was notable then?) started in the WikiTree G2G support forum.

Since WikiTree is a webtree, the idea was a ‘Namesake’ hyperlinking template. It would let the namesake portion of a name be linked to their eponym… plus give a reason for the honor.