Shaping Latin American Spanish at Google

Spanish is extensively spoken in 21 countries, including the United States, and is an official language in 22. More than 400 million people communicate using this language, even before taking into account the US Latino community, the second-largest Hispanic population worldwide. Thus, the long-standing question from the editorial, traditional media and entertainment fields of how to wisely approach translation into Spanish in the face of synchronic linguistic variation, cultural diversity and market differences, can only be answered in industry and product-specific contexts.

The principal and most interesting challenge of Spanish localization for any international company is the tension between the desire to reach a wide audience and the various centers of the language that have normative validity — what scholars call multipolarity. To create Spanish localization operations that are able to meet our business needs and provide the user experience we intend, we need to clearly discern these focal points of the language. It is also necessary to fully understand how our users speak and the ties between their language and their identity. If we are addressing Spanish-speaking Americas, a market composed of 20 countries, we need to set off with a respectful attitude toward the different modalities that form part of the language and incorporate this diversity into the way we work. Not only will this increase the quality of our products, but it will reduce the risk of linguistic choices that in the best case could alienate the majority of our users and in the worst case could carry inappropriate connotations and be offensive for them.

While many locales can be associated with the dominant linguistic norm of one country, locales like Latin American Spanish (es-419) or US Spanish (es-US) handle and balance several of these norms. es-419 takes its tag from the area code employed by the United Nations Statistics Division to group the geographical region composed of Latin America and the Caribbean. It has evolved in the context of previous successful efforts to increase the acceptance of contents by Spanish-speaking Latin American audiences, such as dubbing in film and television. In this manner, es-419 is affirmative of the richness of the language, as it is able to focus on linguistic communities across the Americas, while it strives to make the implicit and often elusive concept of a “neutral” Spanish explicit and tangible.

 

Latin American Spanish at Google

Linguistic content interacts with user-centered design to constitute an essential part of Google products and convey the characteristics of the Google brand. Like engineering teams, those of us who manage written content want to facilitate the instant access of relevant information instead of standing in the way. With a genuine conviction on the efficacy of this goal, the localization team is set on helping our users throughout the world do what they want to do in a quicker and simpler way. We want localized content to contribute to the seamless and enjoyable user experience that the rest of the company pursues and we work hard to carefully balance the characteristics of the Google brand with the linguistic and cultural norms in our markets to allow our products to speak directly to users. Making sure the es-419 versions are aligned with this objective is a complex and rewarding undertaking due to the variation that exists in Spanish along the lexical and syntactical axes.

Imagine you opened your refrigerator and poured a glass of juice. If you were a Spanish speaker, depending on your location, you would have opened a refrigerador, a refrigeradora, a heladera, a nevera or a frigorífico and you would have poured either a jugo or a zumo. As every writer or translator who has strived to reach an international audience knows, we first need to understand the distribution of each variation to determine the best approach. As Google Trends and the Corpus de Referencia del Español Actual (CREA) can confirm, speakers in Mexico and Chile use refrigerador. In Central American countries such as Costa Rica, and in the Pacific coast of northern South America, like Peru, the feminine refrigeradora is more common. In the Río de la Plata region — Argentina, for example — speakers use heladera, and in Colombia and Venezuela, which face the Caribbean Sea in northern South America, speakers use nevera. In Spain, frigorífico is the most used variation, followed by nevera. For the drink, jugo is the principal form across the Spanish-speaking Americas while zumo is the principal form in Spain. If we had imagined washing the glass, we would have assessed the variations of the noun sink (pila, pileta and fregadero) and faucet (llave, grifo, canilla, pila and pluma).

Localization professionals dedicated to Spanish know well some instances of geographical or diatopical lexical variation, including equivalents of words such as computer, laptop, tablet, mobile phone, mouse and speakers, but are also constantly mindful of instances that can exceed the scope of terminological databases. At Google, we frequently remind translators who work on es-419 content to use all the references at their disposal to make informed and creative choices that will speak better to our regional user base. We also ask them to consult us if they are uncertain of the best option. The most effective approach is informed by the ability to see the language as a whole, which not only helps us to avoid unnecessary regionalisms, but also guides us toward the centers that disseminate the linguistic characteristics that are more widely accepted across the region.

At Google, es-419 is shaped in close collaboration with stakeholders from our offices in Mexico, Colombia, Peru, Argentina and Chile. We work with marketing teams to make sure that the linguistic guidelines that give form to es-419 serve our large user base and our day-to-day workload involves regular communication across the region to finalize the names of products, features and key terminology. It is not uncommon for us to join forces with marketers or product experts based in at least two offices before making an important localization decision, all while remaining efficient and keeping up with the speed of the company. The collaborative disposition of Googlers has also allowed us to build a network of country ambassadors who were born and have lived in different Spanish-speaking Latin American countries, and this has allowed us to go beyond es-419 and be more locally relevant where it can have more impact.

As celebrations roll throughout the year, several fun and very visible projects are localized. Last year, we wanted to invite users to play in Santa’s Village and follow his journey in Maps. In mid-December, the following promotion ran on our homepage:

“Skydive with Santa, race with reindeer, and parachute presents in Santa’s Village.”

As each country has a separate domain, we were able to launch more relevant translations for our Mexico and Argentina homepages (Figures 1 and 2):

Salta en paracaídas con Santa, corre con los renos y lanza regalos de Navidad.

Saltá en paracaídas con Papá Noel, corré con los renos y lanzá regalos de Navidad.

Both lines can be translated literally as “Jump in parachute with Santa, run with the reindeer and throw gifts of Christmas.” Their differences stem from the forms of the imperative verbs jump (salta and saltá), run (corre and corré) and throw (lanza and lanzá), which I will explain below. They also vary in the name they give to Santa. Next year, Chileans might just hear from el Viejito Pascuero, their name for Santa.

The shape of American Spanish

The success of a localized product largely depends on translating it using the right register. Maps cannot speak to our users just like AdWords and Transparency Reports cannot sound like the Android site. For es-419, establishing the range of formality we use for products did not take place without consideration of the differences among linguistic communities when it comes to modes of address.

The forms of second-person singular in informal register vary noticeably across Spanish-speaking Latin America. A sentence like “You use Gmail” can look very different depending on the origin of the speaker. Here are examples of the present indicative as it is used throughout the region, with pronouns only shown as reference to indicate verbal form:

(Tú) usas Gmail.

(Vos) usás Gmail.

(Vos) usáis Gmail.

All of these sentences would be translated back literally as “You use Gmail,” referring to one person. The first corresponds to tuteo, the use of the pronoun and historically singular verbal forms. Tuteo is practically exclusive through Mexico, Venezuela, the Antilles and Peru. The last two correspond to voseo, the use of the pronoun vos and historically plural verbal forms. In practice, verbal voseo is not always accompanied by the pronoun vos, but in some countries can be accompanied by the pronoun . Therefore, five different pronominal and verbal combinations of the present indicative can be seen.

As with other written manifestations of “neutral” Spanish inside and outside the localization industry, es-419 is characterised by tuteo. This approach is based on the different degrees of acceptability that voseo holds across the region. While voseo is widely extended through Argentina, Uruguay and Paraguay, its relationship with the prestigious norm and the written language is more nuanced in other countries where it is present. In addition, as has been seen, the linguistic panorama of the voseo is hardly clear-cut. Its most common form is the one that combines the pronoun vos with verbal voseo, as is typical in Argentina (vos usás). However, despite Argentinian influence, in some areas of Uruguay is still used along with verbal voseo ( usás). Typical of Chile is the use of along with the verbal voseo form ending in -is ( usáis). To further outline this complexity, from community to community, the presence and the form of voseo differs among present, past, and future tenses and among indicative, subjunctive and imperative modes.

Formal treatment adds to the richness of the second-person singular. Most Spanish speakers use the pronoun usted along with a third-person form to denote distance and respect, as if they were speaking about a person different from the one they are addressing (Usted usa Analytics). The roles of vos, , and usted, however, sometimes vary across communities and coexist in some of them to express different needs. In Colombia, for example, it is not uncommon to use usted to express familiarity.

Despite the distinction between informal and formal modes of address in second-person singular, Spanish speakers in Latin America do not differentiate between them in plural form, like they do in most of Spain, outside western Andalusia and the Canary Islands. In this manner, regardless of the level of familiarity or distance one wishes to convey, the sentence “You use Gmail”, directed to more than one person, would be translated as Ustedes usan Gmail for a Latin American audience. In Spain, however, it could be translated as the informal Vosotros usáis Gmail or the formal Ustedes usan Gmail.

 

Committing to the user base

Google and many other companies are interested in addressing the global Hispanic community. For any multinational business, the opportunity in Spanish-speaking Americas is evident and locales such as es-419 can allow extensive support. The offering of Latin American Spanish would be impossible without the existence of a collective identity that is able to transcend national political boundaries.

At Google, es-419 is committed to its user base, even when this means departing from the prescriptive authority of the Real Academia Española. The wording of its proposed second definition of the noun clic (click) in Figure 3 can serve as a good example: Pulsación que se hace en alguno de los botones del ratón de un ordenador para dar una instrucción tras haber señalado un enlace o icono de la pantalla (literally “Tap that is made in one of the buttons of the mouse of a computer to give an instruction after pointing to a link or icon of the screen”).

While the description might be accurate, the choice of words does not reflect American Spanish. The variation ordenador (computer) is only used in Spain (computadora is the most used variation across Latin America and computador is used in Chile and Colombia) and the Academia itself has identified ratón (mouse) as a form used in Spain. Further, if this entry had an es-419 version, vínculo would have been a better choice than enlace (link) and ícono would have been used instead of icono (icon).

As we have seen, there are instances of lexical and syntactical variation between Iberian Spanish and American Spanish, but there are also differences within the latter. Amidst this diversity, es-419 is shaped to match the regional linguistic standard that has been defined by mass media, now also digital, which speakers from different countries will recognize as “neutral.” To materialize this linguistic consensus, we are guided by a descriptive approach toward the language that moves us to assimilate and reflect usage at a regional scale. The tag es-419 is carried out proudly, as it can only exist as the sum of a real agreement among speakers and a genuine regional collaborative effort.