The Third Translation Revolution

Learn how Spanish translation has evolved since the advent of artificial intelligence.

Varios idiomas (2023)Fundación Antonio de Nebrija

From the advent of language and the emergence of a huge variety of languages up to some 5,000 years ago, translation had always been an oral task.

Aprendiendo nuevos idiomas (2023)Fundación Antonio de Nebrija

Translating between languages is a difficult task, and translating well between languages is very difficult. This is something that humankind has been doing for tens and thousands of years. Until a few decades ago, translation was an exclusively human task.

There have been three major revolutions in written translation: the invention of writing, the advent of machine translation, and the development of artificial intelligence.

Cuneiform inscription of a message from the Persian King Xerxes (2009) by Bjørn Christian TørrissenFundación Antonio de Nebrija

First Revolution: the invention of writing

The invention of writing was a huge cultural revolution. With the first known scripts, such as cuneiform, we can already see multilingual inscriptions.

Inscripción de Behistún (2015) by Hamidreza Sorouri / Persian Dutch NetworkFundación Antonio de Nebrija

A sample: The Behistun Inscription, in Kermanshah (Iran)

We can find the same text written in three different languages: Old Persian, Elamite, and Babylonian. This cuneiform script is used in all three cases, making this inscription the key to deciphering said script.

Old computer from the 70s (2023)Fundación Antonio de Nebrija

Second Revolution: machine translation

Machine translation was one of the practical applications for which the first computers were used back in the 1960s.

Basic translation (2023)Fundación Antonio de Nebrija

The linguistic approach was very simplistic and the computational power extremely limited. Translating between two languages was thought to consist simply of changing the words of one language to their corresponding ones in another.

Russian - English Dictionary (2010) by Victor KorniyenkoFundación Antonio de Nebrija

They soon realized it wasn't that simple. The first attempted machine translations were from English to Russian, but applying this methodology to such diverse languages only resulted in unnatural sentences.

Translation correction (2023)Fundación Antonio de Nebrija

In the case of languages from the same linguistic family, such as Romance languages or Germanic languages, the quality of the translation was better. They just needed a quick proofread, which was much more agile than translating the text from scratch.

Fotografía Inteligencia Artificial con mano humana (2021) by Tara WinsteaOriginal Source: http://pexels.com/

Third Revolution: the advent of Artificial Intelligence

Although it seems very recent, there has been research into artificial intelligence ever since the 1980s.

Nuevos lenguajes (2023)Fundación Antonio de Nebrija

However, major achievements didn't come about until 2010 with the arrival of computers with a computational power a thousand times greater than that of 30 years ago, leading to the incredible boom in the availability of data in digital format (what we call big data).

The third translation revolution (2023)Fundación Antonio de Nebrija

The third revolution: our present

Translation today is governed by neural machine translation (NMT), which has far surpassed the capabilities of the technologies applied up to now, such as rule-based machine translation (RBMT) and statistical machine translation (SMT).

Neural Machine Translation (NMT) (2023)Fundación Antonio de Nebrija

NMT stems from the creation and development of computer-based neural networks, which are trained with bilingual corpora with which they can translate texts with a quality never seen before from a machine.

AI in translation (2023)Fundación Antonio de Nebrija

With this technology and its application to translation processes, what is the role of translators?


We asked Adriana M. Blas, President of the Spanish Association of Translators, Proofreaders, and Interpreters (ASETRAD).

Adriana M. Blas - Presidenta de ASETRAD (2023)Original Source: Asociación Española de Traductores, Correctores e Intérpretes

Adriana M. Blas - ¿Qué papel juega la proximidad geográfica en la traducción?
00:00

What role does geographical proximity play in translation?

Although it may seem that the low geographical dispersion of the Spanish language is a disadvantage, in the field of translation it is an opportunity, as it offers a very broad market for professionals in this industry.

Adriana M. Blas - ¿Qué tipo de documentos se traducen al español?
00:00

What kind of documents are translated into Spanish?

It's often thought that translation is only necessary for cultural or advertising products, but many types of media are translated in the Spanish-speaking market, such as patient information leaflets, lists of ingredients, instructions, agreements between companies, and so on.

Adriana M. Blas - ¿En qué consiste la profesión de traductor?
00:00

What does the profession of translator involve?

A translation professional is someone who makes communication easier, builds bridges between languages and cultures, and uses these bridges to convey knowledge and make it accessible in different languages.

Adriana M. Blas - ¿Cómo se integran la tecnología y la IA en esta profesión?
00:00

How are technology and AI being integrated into this profession?

We use digital tools that streamline translation and speed up work, and we can also implement them to guarantee the quality of the translation. These include virtual dictionaries, glossaries, style guides, text processors, translation memories, etc.

Adriana M. Blas - ¿Qué falta para que la IA sea un aliado en la traducción?
00:00

What will it take for AI to become an ally in translation?

If we want AI to work in our favor, we need to feed the databases of these engines with a good lexicon, a rich vocabulary, and correct grammatical structures typical of Spanish.

Adriana M. Blas - ¿Qué retos enfrenta la IA como herramienta de traducción?
00:00

What are the challenges facing AI as a translation tool?

Some of the challenges AI will have to face in the coming years to position itself as a good translation tool are specialization in technical sectors, understanding nuances and cultural references, and avoiding biases.

Adriana M. Blas - ¿Podrá sustituir la IA al ser humano en la traducción?
00:00

Will AI be able to replace humans in translation?

The human is and will always be necessary in translation. The translation professional will have the criteria to decide on the quality and accuracy of the text being translated.

Adriana M. Blas - ¿Es la traducción una herramienta necesaria?
00:00

Is translation a necessary tool?

Translation is a necessary tool today. It helps us receive knowledge and culture from other countries, as well as taking Spanish output to all corners of the globe. And as for translators, we're still necessary.

Credits: Story

We are grateful for the collaboration and information provided by the Spanish Association of Translators, Proofreaders and Interpreters, especially to its president, Adriana M. Blas, for granting us an interview.


Other studies consulted: El desembarco de la inteligencia artificial en la traducción automática: la revolución que ya está aquí


Content curation and script: Aina Arbona
Graphic Editor: Carmen García
Consulting and review: Lola Pons (Universidad de Sevilla)

Credits: All media
The story featured may in some cases have been created by an independent third party and may not always represent the views of the institutions, listed below, who have supplied the content.
Explore more
Related theme
Spanish: Beyond the Words
Explore the culture, communities, and stories behind the Spanish language
View theme

Interested in Performance?

Get updates with your personalized Culture Weekly

You are all set!

Your first Culture Weekly will arrive this week.

Home
Discover
Play
Nearby
Favorites