What is new about Slow Translation Movement?

Yes, the movement that brings back the art of doing things by hand, as a form of rebellion against the fast pace of technology and capital, has reached the translation sector.

The Institute of Translation and Interpreting proposed the idea of "Slow Translation Movement" on its website, inspired by other movements with the same “vibe” or “essence”: Slow Food, Slow Fashion, Slow Living, etc.

Applied to the reality of language services, it argues that translation is a job that should be done carefully, with research, time, precision, skill, and artistic talent.

In support of the movement, AI tools should be used very cautiously, in less risky, quicker translations, but the use of technology shouldn't be taken for granted as it is today.

These arguments seem reasonable, but let’s surf over superficiality to understand what’s new about this slow translation movement, and what applies to modern reality.

Understanding the Principles of the Slow Translation Movement

The Institute of Translation and Interpreting, which is a professional association of translators, interpreters, and language service companies in the UK, affiliated with the International Federation of Translators, views the slow translation approach based on the following principles:

1. Celebrate Linguistic Diversity:

Languages carry history and culture, so diversity should be celebrated (and preserved) in any translation. A critique of excessive literal translations that neglect cultural context and richness.

2. Prioritize Art Over Automation:

When the word art appears in the context of translation, it usually refers to the “human touch” and its irreplaceable value. A critique of automation that disregards human experience and expertise.

3. Defend Quality:

Is there more quality in slower translations? For the Slow Translation Movement, yes. Rushed translations can lead to mistakes. More polished and reviewed translations tend to be of higher quality.

4. Maintain High Ethical Standards:

Ethics is always an important issue in the translation world. For the Slow Translation Movement, only the human translator carries the responsibility of conveying the message accurately and with respect for the original language.

5. Demonstrate Intercultural Competence:

Perhaps the most repeated argument for artisanal translation is about cultural competence. The movement argues that only qualified translators have a deep understanding of the cultural dimension of language, and this competence should be valued, including with appropriate financial compensation.

Reflecting and Balancing on the Principles of the Slow Translation Movement

The principles of the Slow Translation Movement mentioned above are all fair, and it seems coherent to keep them in mind for any translation project.

After all, who wouldn't want a project with quality, ethics, linguistic diversity, faithfulness to the language pair, and high standards? 

Any client will want to embrace these attributes, as long as they also come with performance results, of course. 

Let's make an overview reflecting each principle.

• Linguistic Diversity:

Language is culture and history; this is undeniable. Just as it's undeniable that language is a code. Both definitions don't cancel each other out. On the contrary, they complement each other.For machine learning, it's clear that the interpretation is of the code, the data. It's also true that much sensitivity is lost, even though AI is getting better at learning these nuances.Human creative talent cannot be replaced by artificial intelligence, no matter how well-trained it is. Differentiated understanding, sensitivity, and human insights should indeed be prioritized, and these skills are still at play when we talk about Augmented Translation. They are not simply discarded or put in the background.

• Translation as an Art:

What happens when we decide, for example, to translate a project using a CAT Tool? The bulk of the translation is done by machine translation, using integrated tools like translation memory (which identifies the translator's style patterns) and glossary (suggesting words appropriate for the project).The refinement, or the “human touch” – as some like to call it – remains a priority. It's the human translator who will make the choices about which suggestions to accept, and which to reject. There’s no role overlap, maybe just a division of tasks, leaving the human translator with the final say, like the architect of the text.Art doesn't disappear when we use technology. It continues to be part of the translation process, maybe not as “baroque style” as expected by the more conservationist, but as a contemporary art-translation.

• Quality:

Quality goes hand in hand with all these ethical and linguistic principles. What changes in times of technological inclusion in translation processes is that it can involve a performance result that goes beyond linguistic standards.What does this mean? Imagine a translation project for a company’s blog or website. A well-written text isn't necessarily a well-indexed text, meaning it doesn’t necessarily have a better chance of appearing in Google searches.And you’d agree that a client requesting a website translation service would want both, right? Linguistic quality, localization efficiency, and ranking. After all, they want the content to communicate and reach people.In this sense, what the use of technology does is change the parameters of quality definition. The result is even more work for the translator, who is now required to gain skills that go beyond linguistic knowledge. And this also affects the budget, of course.

• Ethics:

Ethics is a dilemma rooted in the debate over AI use because we don't always know if the databases used in machine learning were accessed with permission. This might be the most sensitive point.Artificial intelligence is trained with countless pieces of information, data, data... and is there always permission to access this information?For example, chatbots are trained with various sources like books, websites, and scientific research. If we ask ChatGPT where the sources come from, it would, obviously, respond that only legally accessible content is used.

But when it comes to the work of the linguist, ethics isn't lost when the professional uses digital tools and translation apps.

Prioritizing ethics and professionalism should continue to be a guiding light for translators. Confidentiality of content, faithfulness to the original author, respect for the client, clarity about the tools used, deadlines, and fees. All of this involves an ethical stance in translation. These remain the responsibility of the translator.

• Cultural Competence:

Specialization, study, education, and the cultural background of the translator continue to be valued in translation workflow projects.

Automation wasn't created so that people with no knowledge of the language could take on tasks they were not qualified to do. This is a very important point that must always be emphasized.

I personally wouldn’t be able to translate a German-to-Arabic text with quality, no matter how good the technology at my disposal is. Because I have no skills in any languages.

But talking about a competitive market, with AI, some people accept the service for lower prices, even without the perfect skills. So how does this affect the translator's pay? 

Well, certainly the price-value relationship needs to be adjusted for translation projects, just as in other industries.

Professionals more willing to learn to use technology tend to take on jobs from those who don’t show the same flexibility.

Technology use indeed changes pricing. If you take half the time you usually take on a project, why not charge less? But if the project is more complex, why not charge more?

Even the idea of charging per word becomes outdated in this context.

It seems logical that the more valuable the professional (experience, education, investment in software, cultural immersion in the language), the higher the price they can charge for their services, always considering the complexity required by the project.

Concluding: Better Saying, Provoking

Giving credit to the proposed reflection

Bringing the ideas of the Slow Translation Movement here is more an opportunity to discuss the foundations than a need to persuade you to join or reject the movement.

When analyzing the guidelines, what the movement advocates are nothing more than relevant demands for quality translation projects: ethics, competence, and valuing the human translator.

The critique of all slow movements, like slow food, for example, is about the acceleration of our times. "Speed became our shackles," as the Slow Food Movement states. 

And in this regard, they have a fair point. The pressure for greater productivity in what is often an inhumane amount of time is sickening, and there are already scientific studies warning about this.

What could be a proposal?

We know that the configuration of the modern world is fast. But a bold proposal would be: Couldn’t productivity, instead of always being linked to haste and pressure, be better associated with efficiency?

This way, efficiency could satisfy both the client and the linguist/translator’s quality principles. Not abandoning the delivery of results, but not delivering just anything.

It seems like a fair balance, but it demands skill from the professional. And it’s a bold proposal because it requires work to put things in the right place.

Even with all the agreements, the issue seems to remain the same: the use of AI.

The Slow Translation Movement doesn't express that digital tools should be abandoned but that they should be used in moderation.

And in this regard, they’re correct. The use of artificial intelligence without purpose is as useless as it is burdensome.

In the end, the Slow Translation Movement has brought a good reflection on where we want to take the translation industry in the future. 

Drawing a parallel with the slow food movement, the “flavor” of translation projects will depend on the human translator's sensibility and skills (the fine cook here) in using the right ingredients.