Ellen Singer isn’t afraid to take on new challenges. After cofounding a translation agency and expanding her skills to include copywriting and programming, she’s now most interested in sharing her knowledge through teaching and public speaking. Here, she shares her advice for how to thrive as a linguist in an ever-changing landscape.

How did you get involved in the translation business?

When I was two years old, my Dutch parents moved to Argentina and eventually Uruguay. Dutch was spoken at home, but outside our home we spoke English and Spanish. I grew up trilingual, but perhaps multilingual describes me better, as I believe that math and programming are also languages we can and should master.

In 1995, I was unimpressed by texts and translations around me and decided we could do better. Our agency was founded by four people; all of us had studied programming at some level, and we were at a career crossroad then. The idea was to not only translate and localize, but also program and provide high-quality combined services. One colleague created our first website, which was also the first website for translation services in the Netherlands. We started translating information technology (IT) and other technical documentation; we then moved to the marketing of technology as many marketing copywriters and translators did not understand computer jargon. With time, two co-owners decided to work elsewhere, leaving my husband and me in charge.

We have worked based on the four-eye principle from the start. We believe everyone makes mistakes, but a second pair of eyes — a second translator — can spot most issues. We also believe that the more diverse our team is, in each and every sense of the word, the better.

I have specializations, but I also need challenges, so at times, I take on projects that are outside my comfort zone. I learn something every day. Curiosity killed the cat, but it keeps me motivated. I love meeting multicultural people, because they know there’s more than one way to skin a cat. This idiom is not to be taken literally, especially not by a translator as we depend on our computer-assisted technology (CAT) tools.

Since you entered the translation industry, how has the business landscape changed?

Business is never static. What once worked stops working. What works now will no longer work sometime in the future. You need to constantly learn, change, and adapt. This means reinventing your processes, adjusting your niche, and building on your skills. Standing still isn’t an option.

When we founded AzTech Solutions, we would receive texts by fax or on floppy disks sent by post. It all took time and effort. Now, receipt is almost instantaneous, so clients tend to expect their translations faster. They increasingly expect the translation process to be as fast as lightning. It can be, but this decreases the quality of the work, which is where humans can make a difference.

In our early days, most translation work was assigned by agencies. We would receive the text and translate it into our native languages. We worked from English, Spanish, and Dutch into these same languages and often accepted the same texts into other European languages, which we outsourced to trusted colleagues. Nowadays, we mostly work for direct clients.

In the 90s, translators were often asked to interpret as well. These two tasks are now generally performed by specialists in translation or in interpretation. Another change relates to specialization. Linguists now specialize in niches, although I hear more about diversification since Covid-19. As I grow older, I think everything is sinusoidal, not progressive.

Could you share your experience working with your first client or on your first project?

My first project was in 1995. I can’t remember what it was about, but it was for an agency that provided translations of technical manuals in English. These translations were written by Japanese technicians. It was our job to check and edit them.

I remember that they sent us the same sentence for translation seven times. Each time, we delivered the same translation. They eventually called us because they didn’t believe we were translating accurately. We weren’t! From the first, we had understood what should have been written and translated instead of trying to reflect the actual sentence. The first version was “Pull cord to prevent phire of amplifier.” I don’t remember the final version, but it included the word not, so there was quite a big change. 

This same agency soon provided an early CAT tool, along with a two-day training course. They offered training because many translators were willing but unable to work with the tool. When they first sent the tool and a project, we noticed that the agency’s instructions lacked detail. When they called to ask if we had managed to install the program, we told them that we had not. As other language teams had not been able to do so, we installed the program, edited the instructions, and sent these to them. Our instructions worked and everyone could deliver the project on time. This ensured they respected us!

Providing added value has been a key service component from the very beginning. We point out mistakes in the source text, ask questions when the text isn’t clear, and make any other contributions we think are helpful. We still have clients today who have been entrusting their translations to us from the very beginning.

Do you believe it’s a good time to enter the translation business?

At any given time, there are translators who struggle and those who have enough work. The required skills change over time, and you need to be adaptable. With the advent of CAT tools, many translators decided to retire because they wanted to keep working on paper. I liked the CAT tools. They meant you didn’t have to find and copy previous translations and then make sure you made the appropriate changes.

Integrating useful tools into your work processes is paramount, but every translator has to decide for themselves what they find useful. I find that machine translation (MT) and artificial intelligence (AI) translations don’t yet provide a readable text. Their output seems to be reasonable until you read it closely. Human translators are probably some of the few people who read a complete manual. This puts them in a unique position to spot mistakes at a macro level and realize that a sentence can be interpreted in multiple ways.

The language services market is growing. Some translation work can be done by machines or AI. The output is improving, and we should keep an eye on developments. But to achieve quality output, the engine must be up to the job and trained on relevant data and the company’s existing translations. At TEF 2024, I learned that if AI translates based on a large language model (LLM) that includes a “smaller” language such as Swahili complemented by English data, the quality of the Swahili texts improves. According to the presenter, it is because AI was trained on the universal language aspects of the English. I will keep an eye out for more information.

Obsolete technology is put out to pasture. The same goes for translators who translate like machines and who let their skills become outdated. Translators who can write, who add value, and who make texts resonate with the target audience will always be in demand. But even then, you will only have enough work when you market yourself, when you network, and when people know you exist.

Where do you see yourself professionally in the next 10 years?

I never look that far ahead. I let opportunity knock and let it in. Years ago, I went to a conference and then another. I realized the speakers often had less experience on the topic than I did and that they left out relevant details. I took this opportunity to go to more conferences, but as a speaker. 

One day, the telephone rang and I was asked if I wanted to teach languages. I said yes, and I have been a teacher for many years now. I get bored easily. So, when someone knocks on the door with an opportunity to re-invent myself, I take it. This sometimes means I end up taking on projects that require different skills. For example, carrying out technical translations and teaching two languages to ensure the company employees understand each other. I teach based on company material, often writing or improving their material along the way. Teaching the required jargon is a challenge I gladly take on.

I am getting on in years. I might work fewer hours in 10 years’ time. My children are adults now. They don’t need me much and aren’t as financially dependent on my partner and me.

What predictions do you have for the future of the translation industry?

I don’t call it an industry. Industries produce commodities. We can’t produce texts on a conveyor belt. No two translators would ever produce the same “product.” Writing is an art, a skill. As long as you create quality texts, you will have work. Linguists will also be asked to play roles that don’t exist yet.

When Google Translate (GT) was launched, Google took all the content on multilingual websites on the web and added that content to its data corpus. Many of these sites had been produced by GT. This meant that any mistakes in these texts compounded the issues in GT texts and so its quality decreased. Google then hired linguists to assess the quality of its translations. Nobody could have foreseen this. Just like nobody realized that the advent of the mobile phone would decrease chewing gum sales. Predicting the future is a hit-and-miss activity. I prefer to adapt to actual changes and am always ready for the next challenge.