Pages

Tuesday, February 11, 2020

3 Ways You Can Become an ‘Augmented Translator’

This is a guest post by 
 2019 United Nations Conference on Trade and Development Digital Economy Report, which shows that global internet protocol (IP) traffic, a proxy for data flows, grew from about 100 gigabytes (GB) per day in 1992 to more than 45,000 GB per second in 2017. By 2022, the figure is expected to stand at 150,700 GB per second.





======


The future of professional translation is here. Are you ready? Translation is driving the globalization of communication, but it encompasses more than just translation: linguistic advising, review, proofreading, transcreation, subtitling, language consultancy, linguistic content management... the list goes on. No doubt 2020 and beyond is set to increase opportunities for translators to add value to their clients. But, given the rapidly changing world we now live in, how can translators evolve their own services, becoming 'Augmented Translators'?


Engage with technology


The purpose of technology in translation has always been to help translators deliver and finalize content faster. The days when translators were locked up in a library with a pile of dictionaries and a pencil to produce a translation are long gone. Today, content is processed online, from brochures and web pages to user manuals and market outlooks. Even traditional white papers are no longer exclusively published in paper format. And the list of tools, plug-ins, and technologies available to help translators to finalize and reach audiences continues to grow: translation memories, terminology databases, fragment matches, upLIFT, Neural Machine translation, Autosuggest dictionaries, and more.

Even our corporate language is changing with technology: instead of “engaging” with customers, companies “connect” with customers. Now, for those who are familiar with the technologies offered by our flagship solution, SDL Trados Studio, check out the many assumptions raised about our future by language specialists here.

Even more tech-savvy? Check the other side of the fence and see how content will impact the augmented translators’ environment. Discover SDL Content Assistant, a technology that was considered science-fiction several years ago – but is now very real.

Also, with today’s technology, the help provided to translators does not only come from the tools: now, even content creates itself. Now, it’s up to us, translators, to transform it for our local audience.


Specialize in quality levels, not only in specific industries


Fact: the amount of content to translate has reached incredible levels. While SDL translates hundreds of billions of words every year, this figure remains a drop in an ocean of all translated words. What matters is not the amount to translate. What matters is that the result displayed to your audience meets the quality level expected for such content.

However, billions of words also mean billions of possibilities, and augmented translators are aware of one truth that is the current state of affairs: there is no “standard translation”.

All translations are unique, because clients have unique needs, like their customers. And they also have unique constraints, terminologies, processes, and practices.

With the client’s needs in mind, the augmented translator will adjust their effort and the amount of time required to complete their tasks. And the productivity tools available nowadays are here to help them alleviate the burden: the augmented translator never translates from scratch.

The key factor here is to find the perfect dosage in productivity, the right balance between effort and result. It is important to have a strong understanding of the translation workflow, the tools and assets at your disposal, and your own strengths and skills. This will help assess the quality and thus reduce risks.

In fact, “quality” can only be assessed by a human mind, and this is where the augmented translator and the client can collaborate to set expectations on quality. Because both clients and translators know that a “lack of quality” also means “rework”. And while a “high-quality translation” may be expensive, a “low-quality translation” may cost even more.


Inject culture, and acquire knowledge


Augmented translators will speed up the process of integrating their clients’ requirements to get the quality needed, and that is a truth for all industries. But only if they have adequate assets to help them get started in an augmented world.

An augmented translator will take advantage of the following resources:
  • Content reuse from translation memories
  • Glossaries to apply preferred terminology
  • Style guides to comply with formatting, grammar, and stylistic rules 
  • The tone of voice or brand guides to convey the brand’s message 
  • Project-specific instructions, like character limitations
  • Machine and AI-enabled translation engines to accelerate productivity
All these automated tools and assets are literally “knowledge providers” to the translator, and help non-specialized translators to meet client requests even without even knowing the client. These knowledge providers are useful since all the clients have preferred terms, favorite wordings, and different rules.

Of course, this automation can also be error-prone and full of traps: terms in glossaries that do not take the context into account, incorrect source texts written by non-native speakers, corporate jargon not understandable outside of your client’s professional sphere, and more.

This is where the augmented translator has two strong cards to play: culture and understanding.

Augmented translators will be able to spot errors in the source text, avoid using offensive or restrictive content, use the appropriate language for the target audience, rewrite puns, detect dual meanings, adapt to stylistic rules, and correct erroneous terminology used by translation engines, etc.

The augmented translator walks in the footsteps of the ancient copyists and scribes and embraces the same mission and ambition: connect cultures and content to share a message.


*****




Jonathan Grisot started as a translator in 2007 and currently holds a position of Senior Language Specialist for SDL in Paris. He is responsible for driving Machine Translation initiatives, managing internal training and quality best practices and is still involved in various translation and transcreation projects. He is also managing the Junior Academy, a local SDL onboarding structure for newly hired SDL translators. Born in Burgundy and raised on the French Riviera, Jonathan considers his detective novels, sci-fi and fantasy books as his numerous children.


4 comments:

  1. Great text! I'm sharing it with people at Translators101.
    Thanks Kirti.
    Thanks Jonathan.

    ReplyDelete
    Replies
    1. Jonathan GrisotMay 4, 2020 at 7:24 AM

      Hi William, and thank you for your message!

      Delete
  2. I find the following quote quite curious:
    "Augmented translators will be able to spot errors in the source text, avoid using offensive or restrictive content, use the appropriate language for the target audience, rewrite puns, detect dual meanings, adapt to stylistic rules, and correct erroneous terminology used by translation engines, etc."

    Well... This is exactly what translators already do. This is a summary of the core and real skills and knowledge required to produce a good translation, alongside the prerequisite the "translation industry" likes to gloss over as if it it nothing: the ability to write. Something many people claim they have but fewer actually do. Translation is a piece of writing, not a mechanical conversion from language to another. Any piece of text that required effort, attention, skill, and mindful crafting requires the same when translated.

    Furthermore, as the recent Linguist Supply Chain survey (i.e. the agency supply chain) by CSA (https://insights.csa-research.com/reportaction/305013106/Toc) concluded: "On average, linguists earn US$29,000 per year before taxes. One in five respondents (21%) earn less than US$5,000 annually, making translation a source of supplemental income more than a career – as was confirmed earlier in the career focus findings."

    I wonder if it will be mentioned to buyers that all those augmented goodies that are claimed to -- all the skillful and mindful adaptations -- will be provided by glorified amateurs who are unlikely to have the focus, competence, and dedication to carry out all of this to a reasonably high level.

    As hinted in the article itself: "All these automated tools and assets are literally “knowledge providers” to the translator, and help non-specialized translators to meet client requests…"
    A non-specialized translator is not a translator at all, just as as a civil engineer is not an aerospace engineer. If you don't know the field and what you're a doing, then your umbrella title is meaningless. There are already plenty of amateur translators who google their way through all of their translations and the results pretty much are what one could expect. Machine Translation is not a magic bullet solution for bridging fundamental knowledge and skill gaps.
    Why would you even consider hiring a non-specialist to do what one clearly acknowledges is a specialized service (except for cutting costs by hiring casual amateurs)?

    ReplyDelete
    Replies
    1. Jonathan GrisotMay 4, 2020 at 7:24 AM

      Hi Shai, and thank you for your extensive reply, which allows me to explain my views.
      Regarding “What translators already do”: translators do these things indeed, but they now have to do it faster as the volume of data grows. And to complete more and more work, they need assistants and a solid tech backbone, which is NMT as I write.
      It is like explaining that you’d better take your car or the train or plane to travel from one country to another, and someone in the crowd says that “we already do that” with a horse carriage. The image is a bit rough, but quite self-explanatory.
      Translators expressed the same doubts and concerns when the translation memories were created: translation memories store our translations that will be reused, we will no longer have work as everything will already come from the memories, etc., and look what happened: companies create more products and need more supporting material to localize, TMs are not applicable to all clients who have very specific jargon, new content needs translation.
      And while these translation memories are now part of our daily work, we still translate!
      Now, on the ability to write, I agree with you, but to a certain extent. The ability to master your language is one core skill, the ability to meet the clients’ expectations is now even more crucial.
      There are countless examples of wrong formulas, weak wording and more-than-unusual client preferences that show how unimportant our ability to write is. Professional translators are very sad when their beautiful piece of translation altered, trampled and mixed in to chunks of jargon by a client, but this happens more and more.
      Several clients value this ability to write and value translators that go beyond the simple translation process. But many content providers need “publishable” content, not “nice” content (even if they say the contrary).
      On “non-specialized translators”, what I mean is that a translator specialized in, say, automotive content can also deal with standard IT content or even standard medical content (like troubleshooting procedures or a manual for a syringe pump), if they have a robust MT engine, clean and extensive translations memories and glossaries.
      Provided that the translator is professional and is doing is homework correctly. Because technology and machine translation shall not prevent us to do some research. You say that “Machine Translation is not a magic bullet solution for bridging fundamental knowledge”: it does not entirely, but it can, for some content. We can view NMT and ling tech as a rope bridge, where company buyers already expect a concrete bridge able to process their wagons of content. But still, there is a way to cross the detroit.
      Finally, “non-specialized” does not mean “amateur” to me. This blog was not intended to people who consider or practice translation as a casual activity. But I meant to explain that professional translators could also use new technologies to extend, step by step, their area of specialization and their productivity as well.

      Delete