Interpreting's Awkward Adolescence: Growing into a World of Possibilities
By Barry Slaughter Olsen and Katharine Allen, InterpretAmerica
Growing pains. We’ve all had them. They come at that awkward time known as adolescence, when youngsters begin to stretch and grow as they discover a new, more expansive world full of possibilities and challenges.
Interpreting is going through its own growing pains. Revenue generated from the language services industry is now bigger than 40% of the world’s countries, and demand continues to grow. (The Language Services Market 2014, Common Sense Advisory, Inc.) In interpreting, demand currently outstrips supply as suppliers struggle to meet the need for interpreting services in new places and new ways.
Four key trends in the language market present significant challenges for companies working to scale their service offerings to meet growing demand:
- The growth in content.
- More languages coming onto the global market needing interpreting.
- The desire for “anytime” language service;
- The desire for “anywhere” language service.
In other words, meeting global demand for translation and interpreting is increasingly dependent on how services are scaled up. Translation services are almost exclusively asynchronous. When global demand exploded for written content, machine translation is what ultimately made it possible for the once small, cottage industry to blossom into a major, global service industry. This transition is far from over, but translation is a good decade farther down the path when compared to interpreting.
Interpreting services facilitate synchronous communication. Worldwide penetration of mobile devices is the spark now leading to a similar explosion in the demand for real-time telephonic, chat, and video communication. How will it be possible to scale up a live, labor-intensive service when interpreted interactions may last anywhere from 5 minutes to 5 hours?
If Google, Yahoo, Microsoft and a host of other internet companies have their way, the answer will be: speech-to-speech translation, or what many practitioners and laypeople refer to as machine interpreting. Microsoft made a high-profile splash earlier this year when it showed off Skype Translator, “an upcoming feature aiming at real-time interpretation of voice calls within two years.” Mind you, this is “real-time interpretation” without any involvement of a “real-live interpreter.”
We would submit, however, that this “real-time interpretation” moniker is misleading. Machines do not interpret anything. They crunch numbers and words, they collect and store data and they recognize complex patterns. Whatever the machines ultimately produce, the interpretation of what it all actually means is still left up to humans. And this is definitely the case with speech-to-speech translation as well. But make no mistake, these technologies are going to be sold to the general public as a low-cost real-time replacement for interpreters. What this will ultimately mean for the language industry as a whole and for working interpreters is still unclear.
Perhaps the answer will lie in a hybrid approach leveraging real-time human interaction with machine processing. Just this week, Stanford University researchers announced progress on a program that combines software with human intelligence to improve translation. With the improvement and spread of Siri-like voice over technologies, it is not hard to imagine similar hybrid solutions for interpreting.
In a final example, Australia’s Victorian Interpreting and Translating Service (VITS) has leveraged old school technology to scale services for multiple language groups across a vast territory to facilitate access to voting. Its service provides automated, recorded voice answers to simple phone queries in multiple languages that will nonetheless take the caller to a live interpreter when needed. This approach saves human interpreter services for those interactions where they are really needed, helping to address the worldwide gap in the number of trained interpreters compared to demand.
With the significant headway and the abundant hype surrounding the automation of translation and interpreting, a recent “Proposer’s Day” announcement from the Defense Advanced Research Projects Agency (DARPA) focusing specifically on innovating with new technological solutions for communicating across langauges is yet another example of the high-level interest in scaling up interpreting services in entirely new ways.
Next month, Katharine Allen and Barry S. Olsen, co-presidents of InterpretAmerica, will address this topic for GALA’s webinar series in The Race to "Anytime, Anywhere" Interpreting on 13 November. Next spring, they’ll help facilitate the discussion in more depth at think! Interpreting at GALA’s annual Language of Business conference in Seville, Spain. Come join this critical discussion on the future of our profession.
Barry Slaughter Olsen, is a veteran conference interpreter and technophile with over two decades of experience interpreting, training interpreters and organizing language services. He is the founder and co-president of InterpretAmerica, a professor at the Monterey Institute of International Studies and General Manager of multilingual at ZipDX. Follow Barry @ProfessorOlsen.
Katharine Allen is co-President of InterpretAmerica, an organization dedicated to raising the profile of the interpreting industry, which hosts the annual InterpretAmerica Summit. She trains interpreters and interpreter trainers, provides curriculum design, and language access consulting to hospitals. Clients have included Kaiser Permanente, the Department of Defense, Glendon College of Translation and the Monterey Institute of International Studies (MIIS). Katharine holds a Masters in Translation and Interpretation from MIIS. Follow Katharine @interpamerica.