FactGrid talk:Names: Difference between revisions

From FactGrid
Jump to navigation Jump to search
(Created page with "== Reasons why a merger of given names and family names might make sense == FactGrid follows a development Wikidata has taken. We set items for all personal names and we created item couples if a name was used on both properties P247 and P248 as a family name and as given name. As of May 26 2024 we have 1559 sich item couples of the exact same labels - one to be used as a given name only the other to be used on the family name only The practice is problematic as users...")
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
== Reasons why a merger of given names and family names might make sense ==
== Reasons why a merger of given names and family names might make sense ==
FactGrid follows a development Wikidata has taken. We set items for all personal names and we created item couples if a name was used on both properties P247 and P248 as a family name and as given name.
FactGrid follows more or less the the decisions Wikidata has taken. We created items for all personal names and we created item couples wherever names existed both as family names and as personal given names:


As of May 26 2024 we have 1559 sich item couples of the exact same labels - one to be used as a given name only the other to be used on the family name only
* [[Item:Q31850]] Jacob (given name, prénom, Vorname)
* [[Item:Q146309]] Jacob  (family name, nom de famille, Familienname)


The practice is problematic as users do not always observe the distinction. We could use property constraints to make sure they do it right. One might, however ask whether the entire separation makes sense.
The decision is nice in its interoperability with Wikidata, it is otherwise questionable as it creates a functional and essential gap where we might rather reflect the congruences. As of May 26 2024 we have 1559 such item couples which are not always kept apart by our users.


Many modern family names started as personal names or as genitive-derivates (son/daughter of); today we get personal names from family names in African languages.  
The following is a consideration of why one should (not) put all names into a single category.


There are good reasons to have just one sort of item under a general P2 = Name (Q439355) statement.
=== Names are special ===


The basic functioning is the constant: We do not "translate" names, we "transliterate" them in different writing systems: "Harry Potter" stays Harry Potter in German though the translation should be "Heinrich Töpfer" or Spanish "Henrico Alfarero".  
Names are not translated, they are reproduced in different writing systems, transliterated. The occupation miller is meunier, molinero and Müller in French, Spanish and German. Glen Miller, is, hoverer, in all languages with a Latin script Glen Miller. It changes to Гленн Миллер with the aim to preserve the original sound. (We have name translations in older contexts such as the Bible and medieval and early modern documents - the practice was here that Latin names had different versions in different languages: מַתִּתְיָהוּ, is Greek Μαθθαῖον, Latin Matthaeus, German Matthäus, English Matthew, French Mattheu, or Spanish Mateo to reflect the traditions of nationalised Latin name usage.


Given names have fixed language variants in a certain contexts such as the Bible or medieval and early modern cultural exchanges with variant equivalents such as מַתִּתְיָהוּ, Μαθθαῖον, Matthaeus, Matthäus, Matthew, Mattheu, Mateo.
First names and family names are at the same moment not that different. Many family names are based on first names (or genitives of first given names such as in Adam and Adams). The etymologies are the same. The functionalities can differ as the entire concept of Family names developed differently around the world.


Under a general P2 = Name (Q439355) we could state P280 functions "given name", "male given name", "female given name", "dual sex given name", "family name"
The question is here: Why not express the difference solely with the Property that introduces the name. John the given name follows on [[Property:P248]]. Where the same name follows on [[Property:P247]] it is used as a family name.
 
It is not that clear whether there is an intrinsic difference other than the difference of the position in an name. On Wikibase it is the property that defines the position of the name in a first-name/second-name compound.
 
=== What would happe if we had just one name "Jacob" - sometimes used as given, sometimes as family name? ===
 
* Etymologies should be the same
* Statistical and geographical occurrences are stated with a look at the occurrence of people that have that name as given or family name.
* Legal issues like registered trademarks should be the same on both
* Transliterations are the same
 
=== A unified data model ===
* The standard P2 statement "instance of" should state "Name" ([[Item:Q439355]])
* [[Property:P280]] should state various functions such as "male given name", "female given name", "dual sex given name", "family name"
* [[Property:P625]] could be deprecated for a more complex ontology of usages
* [[Property:P987]] and [[Property:P988]] would state etymological contexts

Revision as of 23:19, 26 May 2024

Reasons why a merger of given names and family names might make sense

FactGrid follows more or less the the decisions Wikidata has taken. We created items for all personal names and we created item couples wherever names existed both as family names and as personal given names:

The decision is nice in its interoperability with Wikidata, it is otherwise questionable as it creates a functional and essential gap where we might rather reflect the congruences. As of May 26 2024 we have 1559 such item couples which are not always kept apart by our users.

The following is a consideration of why one should (not) put all names into a single category.

Names are special

Names are not translated, they are reproduced in different writing systems, transliterated. The occupation miller is meunier, molinero and Müller in French, Spanish and German. Glen Miller, is, hoverer, in all languages with a Latin script Glen Miller. It changes to Гленн Миллер with the aim to preserve the original sound. (We have name translations in older contexts such as the Bible and medieval and early modern documents - the practice was here that Latin names had different versions in different languages: מַתִּתְיָהוּ, is Greek Μαθθαῖον, Latin Matthaeus, German Matthäus, English Matthew, French Mattheu, or Spanish Mateo to reflect the traditions of nationalised Latin name usage.

First names and family names are at the same moment not that different. Many family names are based on first names (or genitives of first given names such as in Adam and Adams). The etymologies are the same. The functionalities can differ as the entire concept of Family names developed differently around the world.

The question is here: Why not express the difference solely with the Property that introduces the name. John the given name follows on Property:P248. Where the same name follows on Property:P247 it is used as a family name.

It is not that clear whether there is an intrinsic difference other than the difference of the position in an name. On Wikibase it is the property that defines the position of the name in a first-name/second-name compound.

What would happe if we had just one name "Jacob" - sometimes used as given, sometimes as family name?

  • Etymologies should be the same
  • Statistical and geographical occurrences are stated with a look at the occurrence of people that have that name as given or family name.
  • Legal issues like registered trademarks should be the same on both
  • Transliterations are the same

A unified data model

  • The standard P2 statement "instance of" should state "Name" (Item:Q439355)
  • Property:P280 should state various functions such as "male given name", "female given name", "dual sex given name", "family name"
  • Property:P625 could be deprecated for a more complex ontology of usages
  • Property:P987 and Property:P988 would state etymological contexts