Tú/Usted in Spanish translations

(Will Thompson) #1

The majority of Spanish speakers are outside Spain. GNOME currently has a single Spanish translation, with very thorough coverage. There’s a placeholder for a Mexican Spanish translation team, but there appear to be no translations there at present, and no team members.

I don’t speak Spanish, so please forgive my ignorance! I understand that there are many differences between different dialects of Spanish, but the one I’m interested in is the choice of second-person pronoun to refer to the user.

In GNOME, Spanish translations use the “usted” form. For example, the “About You” page of Initial Setup is translated as “Acerca de usted”, and “Connect Your Online Accounts” becomes “Conectar a sus cuentas en línea”. In Endless, whose Spanish-speaking users are mainly in Latin America, we have historically modified the Spanish translation downstream to use the “tú” form. For example, “About You” is “Acerca de ti” and “Connect Your Online Accounts” is “Conectar a tus cuentas en línea”.

This change is described internally like this:

any Spanish translations use the familiar “you” (“tu”) rather than formal (“usted” / “su”)

but my understanding from discussion with Spanish speakers and various Wikipedia pages is that it’s a little more complicated than “make the language less formal”, and that referring to the user as “usted” is actually wrong in some dialects of Spanish.

I have several questions:

  1. Is my understanding that “usted” is actually wrong for many Spanish dialects correct?
  2. How do Spanish-speaking GNOME users outside of Spain feel about apps using the “usted” form? How would Spanish-speaking GNOME users in Spain feel about apps using the “tú” form?
  3. If the correct second-person form really is different between Spain and other countries, is there a low-maintenance way to maintain (for example) a “Latin American Spanish” translation that derives from the Spanish translation and only changes the strings that refers to the user? Or would one have to create es_MX, es_GT, es_AR, etc translations?

(Will Thompson) #2

My interest in this is because I have recently become responsible for managing Endless’s downstream translations. While it’s unavoidable that we’ll need to translate new strings that we introduce when we add downstream-only features or changes, I would like to get us out of the habit of adding downstream translations for upstream strings, and (worse) changing existing translations downstream. My predecessor speaks many more languages than I do, so it’s harder for me than it was for him to coordinate these – and maintaining translations upstream is preferable for all the normal reasons.

English has a somewhat similar situation with “colour”/“color” and “organise”/“organize”; answering my three questions as if they were about that:

  1. In the UK, we use “colour” and “organise”; in Canada, “colour” and “organize” are standard; in the US, “color” and “organize” are standard.
  2. GNOME has an en_GB translation, so my desktop uses the spellings from my dialect. It’s very common for third-party apps to not bother with a UK English translation, so I am used to seeing US-dialect spellings, but it bothers me.
  3. I believe the en_GB and en_CA GNOME translations are generated with a script from the en_US source strings. What happens for en_AU and en_IN, where the UK-style spellings are standard but where there is no translation? I don’t know.

(Daniel Mustieles) #3

Hi Will,

Having a separate translation for Spanish (from Spain) and Spanish (latin) would be an ideal scenario for both GNOME and your project since, as you have already said, in some regions of South/Central - America using “usted” is not correct or even not frecuent.

In GNOME Spain we decided to use that term because of is more polite for us, but this is our decision… teams like MX, PE, etc are completely free to take our translations and only modify that term, saving a lot of time and efforts doing so.

In your particular case, if you use upstream translations and have no muscle to translate it into different Spanish variations, I would not worry about using “usted”. AFAIK this is a widely-used term in South/Central - America so there should not be a problem there. If a particular team feels more comfortable using “tú” they should be able to initiate a translation team to change it and distribute the new translations under their own locale (like en_GB currently does, for example… I don’t think they use scripts for doing so, but I don’t know it).

Hope this helps


(Rodrigo Lledó) #4

Hi, I’m half Latin and half Spanish. Also a GNOME translator.
It’s exactly as you said regarding US and UK translations. It bothers some people.
Latin people could see “usted” very cold/distant while Spanish people would like to be treated as customers.
In Ubuntu Spanish translation team we treat the users formally although never using “usted” but “el usuario” (“the user”) because we aim to a global ES translation.
I personally don’t care. In a user guide I don’t mind to be treated formally but in daily use programs is a little bit weird and tiring because I just want to relax and be in an informal register when I’m not working.


(lorabe) #5

I’m mexican but i have never been involved in a work of translation, so these are only my impressions.

I understand the differences between “tú” and “usted” as something cultural, not strictly grammatical. So i can say that it’s not wrong, it’s just more adapted to Spanish people, than to latin american people.

But it is totally understandable. Words that are actually incorrect are “vosotros” “estais” all those words that ends with “ais” because in Latin America we have replaced it. I don’t know if these are in the documentation, but if they are, it would be appropriate to make a replacement.



(Victor Kareh) #6

To be honest, if part of the problem is having to maintain two (or more) variants of Spanish, you could very well default to using the formal and maintain only one.

There are a lot more regional variations than just formal vs familiar forms, so where do you stop then? My suggestion is to stick to the formal version. It is correct in all versions of Spanish, even when not culturally common.

Also, I know plenty of Latin Americans who would not dare tutear (use the familiar tú) people outside their immediate family circle. Conversely, I know plenty of Spaniards that would casually use tú in conversations with strangers and superiors. So it’s not just a Spain vs Latin America difference, even within those there are enough cultural differences to drive you crazy.

By sticking to the formal version, I think would keep a good baseline level of respect for those who expect it, and those who don’t, well, they will not be offended. There are ways of offending people just by using the formal instead of the familiar, but it goes beyond just the choice of words: the context and tone matter.

So tl;dr: using usted would yield the most respect to amount of work ratio, and it has the benefit of being correct and appropriate in all variants of Spanish (that’s why we have the RAE)

NB: I’m from Puerto Rico and would very rarely catch me saying usted to anyone: once I know your name, you become , yet I think usted is more appropriate in this context.


(Victor Kareh) #7

To answer the questions more directly…

No, usted is correct and appropriate in all variants of Spanish. Maybe not common, but certainly correct.

I have no issue with that. Maybe it feels a bit dry, but that’s both my cultural bias and my personal preference.

The lowest maintenance I can think of would be to have “es” as the Latin America version and “es_ES” as the Spain version, but that still assumes that that’s the main difference and that it’s clear-cut between Spain and LatAm


(emma peel) #8

Hello, this topic is very dear for me!

I am a Latin American Spanish native speaker, and some years ago, while translating the Tor Project and Tails, I pushed for the informal second person form when addressing the users. I felt the formal version was not appropriate as it was indicating distance, and in many cases a position of authority, that was not what the original phrase would translate to. Today most of our applications have the informal version.

We have mostly only one Spanish translation around the ecosystem of applications on the Tor Project, although now we are having another group of Argentinian Spanish translators that are very active.

I think it also depends on the application and the context where you are using it, I translate mostly Tails and Tor documentation and interfaces.

I can see that Tails users are a very small subset of the GNOME users, and some other GNOME users may like the formal version, like school directors, heads of government, etc. But for Tails/Tor docs seemed out of place.


(lorabe) #9

I think it’s a good idea to do a general translation for spanish speakers, it would be less tiring to the project.


(Will Thompson) #10

Thanks for all the input!

The other approach I could think of would be to pick a locale to serve as a proxy for “Spanish as spoken in Latin America” and arrange for users who pick a different dialect of Spanish in GNOME Initial Setup to have LANGUAGE set to include both their own locale and this other locale. This is a GNU extension. Then, in that locale, translate only those strings where LatAm usage is different to Spain’s.

As it happens, Endless’ “Spanish” image is es_MX, and there is a Guatemala-specific image with es_GT. So, for a concrete example, GNOME might have an es.po (with the current convention) and es_MX.po that includes only the strings that differ. A Guatemalan user would have LANGUAGE=es_GT:es_MX. Assuming the lookup order is breadth-first not depth-first, gettext would look up each string first in es_GT (which doesn’t exist), then in es_MX (which would exists, but would be a partial translation), then es (which exists, and is complete).

(I haven’t actually tested whether the search is breadth-first or depth-first…)

As a first step, I’ll try to find time to extract our downstream strings and open some MRs adding es_MX.po. This at least would get these changes out of our downstream repos into somewhere where, in principle, something could be done with them. :slight_smile: In particular, this might make it possible for @emmapeel/Tails/Tor to reuse some of this work?