There is some complex discussion about how the names system in GRAMPS can be improved in future versions. There is a bug: http://www.gramps-project.org/bugs/view.php?id=3161 and an email thread: 
This page tries to collect and organise the discussion.
Middle name: http://en.wikipedia.org/wiki/Middle_name
Plural last names: http://en.wikipedia.org/wiki/Middle_name#Spain_and_Latin_America
Family name: http://en.wikipedia.org/wiki/Family_name
Portuguese names: http://en.wikipedia.org/wiki/Portuguese_name
Spanish names: http://en.wikipedia.org/wiki/Spanish_naming_customs
Some relevante quotes from the links above
...In the case of Portuguese naming customs, the main surname (the one used in alphasorting, indexing, abbreviations, and greetings), appears last (reverse the order of Spanish surnames)...
...However, nowadays in Spain and in many Spanish-speaking countries (...) most people have two family names, although in some situations only the first is used. The first family name is the paternal one, inherited from the father's paternal family name. The second family name is the maternal one, inherited from the mother's paternal family name...
...The son of Juan Carlos Pérez Larios and Susana Estela Ríos Domínguez, if given the same first name of his father, would be Juan Carlos Pérez Ríos. Pérez is the "important" last name and the one used if only one is needed...
The reference to a "main surname" and "important last name" is one of the points to have in mind. There are differences between the Spanish and Portuguese naming customs, but this much can be generalised I think:
- The presence of more than one surname, typically taken from both the father and the mother
- The possibility of having more than two surnames, taken from both parents and in some cases from grandparents
- The "main" surname (i.e. the one used in indexing, for example, or the one generally used to denote family affiliation) is not always in a fixed position
The patronymic can be "hard" or "soft", for lack of a better terms, and by this I mean that in medieval times the "surname" was actually a patronymic (Henrique -> Afonso Henriques, etc.), while other times the patronymic is added along with surnames (Rodrigo Garcia -> Gonçalo Rodrigues Garcia). The patronymic is a somewhat different discussion, not directly related to the "multiple surnames" one.
Since existing fields are Given Name and Family Name one would end up, if following the meaning of those words to the letter, to do something like this:
Complete Name: José de Mascarenhas da Silva e Lencastre Given Name: José Family Name: de Mascarenhas da Silva e Lencastre
This is semantically correct, but obscure the real meaning of the name: D. José de Mascarenhas can be said to belong to the Mascarenhas family, and not to the "Mascarenhas da Silva e Lencastre" one.
The general usage I've seen falls within these options:
- Use Given Name for all surnames except the last: this is common in Portugal (and related countries in terms of naming customs) since the last name is nowadays almost always the "main" family name.
- Use Family Name to contain all the surnames: these needs some "group as..." tweaks to work.
- Use some other field in a custom way: generally a bad idea, since consistency is not guaranteed or even expected when using the data for other purposes (reports, etc.)
The discussion is about adding some field (or some other solution) to make the data entry more logical and less likely to need "kludges" in order to appear as expected, both in reports, in the daily use of the program and in terms of expected behaviour (grouping, pre-filling of surnames when adding, etc).
Example names and explanations
- Henri de Blier, son of Collard, sometimes referred to as Henri Collard de Blier
- João Soares de Sousa
- given name(s): João, middle name: Soares (first surname), surname: de Sousa (second surname)
- Joe (the hitter) Silva de Souares
- Given name: Joe, Nickname: The Hitter, Secondary surname: Silva, First surname: Souares, prefix to first surname: de
- Santiago Ramón y Cajal
- García Álvarez de Toledo y Carrillo de Toledo
- José de Mascarenhas da Silva e Lencastre
- given name: José, Family surname: Mascarenhas, prefix to Family surname: de, additional surnames: [da] Silva [e] Lencastre
(every idea we have for now)
The secondary_surname data can be displayed in whatever manner a user wishes by using the proper name format. So each of the following is possible, and controllable by the user.
However, the name display does not effect the GEDCOM output. Someone mentioned a possible method of listing multiple surnames per name in the GEDCOM format.
[Given Names] [Middle Name(s)] [Family Name] Comments: (signed)
[Given name] [Secondary Surnames] [Family Name] Comments: (signed)
Option 1 and 2 are the same, having only the data called something different, right?
[Given Names][Family Name][Secondary Surnames] Comments: (signed)
Table of names
(to be populated once we have a proposed set of headings)