The Democratization of Translation
- Carl Yao
- January 11, 2016
- 3,644 views
As for many of the disruptive technologies to emerge over the last several decades, critics have jumped out of the woodwork to declare why new translation technologies will fall short of their targets.
A recent example of such doubt comes from a piece posted on Numenware, in which the writer, Bob Myers, laments the development of new technologies that try to modernize the age old translation process. He then labels technology innovations in the translation industry as mere “reinventions,” full of “breathless hypes” and “plausible-sounding theories” that would ultimately turn up short. The author cites a number of excerpts from a previous article on Slator, from which the author proceeds to raise the following objections, summarized below:
CSOFT sees huge untapped demand for translation services that can be unlocked only if prices are reduced dramatically by using a crowdsourcing approach. Yao called this the long tail of translation demand…Virtually every phrase in this quote raises questions. Where is this huge untapped demand? From what sector? Who has determined that all these users are price sensitive?
Myers’ arguments are good ones and worth considering. Let’s deconstruct his argument step by step.
The untapped demand in this case refers to the fact that translation today is too expensive or complicated for businesses to make international expansion worthwhile. Admittedly, if you ask companies if they are actively translating today, many of them will likely say no. That’s because under present circumstances, translation isn’t an affordable possibility. Surprisingly, the world has only 250,000 professional translators working in the translation field, yet there are nearly 6,500 languages spoken. The result? We have a huge mismatch between the supply and demand for translation, pushing its price way up.
However, if we come up with a solution that makes translation easy and affordable, we can create a consumer demand for translation that previously didn’t exist. Numenware is right to raise questions about supply and demand, but for Stepes to be truly game-changing (and I believe it is), it will change the actual conditions under which businesses evaluate international expansion. Clients that would never have considered translation before can now afford translation. Large demand doesn’t exist now – or rather, it’s invisible, latent, waiting for the next disruptive innovation to let it spring into being. Good products meet existing unmet demand; the best products fulfill demand users never even knew they had by changing the way we think or behave.
On the translation side, by making translation easily accessible on mobile devices, Stepes will encourage more bilingual people to translate. By 2020, there will be 6.1 billion smartphone users. Even if just a fraction of these users are bilingual, the sheer human translation power we would be able to tap into through mobile translation platforms – as Stepes does – would far exceed the number of professional translators we have today. Being able to capture this much language talent is unprecedented, and it’s understandable to have lots of questions – as do we. However, with the potential to significantly change the way we approach translation and globalization, we believe it’s worth testing out solutions rather than critiquing them.
Next, the author wonders how Stepes and a crowdsourced model would be able make translation service more competitive:
How will crowdsourcing reduce prices? People place a certain value on their time, whether they are translating 50K words per month or 50 words per month. Do they actually understand the actual theory of the long tail, which can be very long but still have a very low area underneath it?
First, the Stepes translation model is able to reduce costs because it cuts out much of the overhead. For example, by automating such tasks as pre-processing, quote creation, progress reporting, manually preparing billing and invoicing, etc., for both the language service provider (LSP) and freelance translators, Stepes is able to effectively reduce the cost per word for translation.
Secondly, Stepes increases productivity gains by simplifying the translation process so translators can better focus on translating instead of being bogged down with translation tags for example. On this same note, by allowing more people with subject matter knowledge to translate texts relevant to their own fields, we increase competition which is usually a good thing for consumer pricing. Stepes also takes advantage of speech to text technology already available on both the iOS and Android phones to input text faster with voice. Combined with machine translation support, Stepes translators are able to translate faster at better quality.
Third, since Stepes uses a crowdsourced translation model that is classified into different industries fields, it’s able to assemble large TM (translation memory) databases that generate translation leverage from previous translation projects, further reducing translation costs to clients.
In the translation and localization business, a small project with a few dozen words or fewer is often regarded as a long tail project; with the traditional translation model, a small project still has to run through the entire, inefficient translation process. However, small translation projects are increasingly becoming commonplace, and by automating or simplifying much of the translation management process, Stepes provides a solution for long tail projects that would normally be a headache within the traditional translation model.
Next, Myers questions the Uber translation model and its ability to deliver quality.
… translation is inextricably related to the subject matter. Unlike Uber which offers three types of cars, it is meaningful to distinguish when translating among hundreds of subject areas, or more. A translation by someone not versed in the subject area is likely to range from unusable, to embarrassing, to damaging.
If anything, a Uber translation model will do better at decreasing the incidence of an all-purpose translator taking on all kind of content, the status quo at most localization companies because of limited numbers of translators.
By making translation more accessible for everyone, Stepes expands the pool of translators available. In addition to supporting professional translators, Stepes also enables subject matter experts – doctors, artists, engineers, lawyers, students, or business managers to name just a few– to translate highly technical content at a much higher quality. Whereas they previously might have been shut out of participating in the translation process, Stepes enables an increasing number of experts to translate in their spare time and earn money while satisfying a desire to contribute to industry knowledge across languages (Interested readers can read more on how the Uber translation models delivers better linguistic quality here). Stepes guarantees quality by carefully classifying its linguists based on specific industry fields, ensuring that only translators with the right subject matter are assigned to work on relevant projects.
Myers goes on to critique Stepes’ chat-based translation process:
Translators on Stepes use a chat-based user interface (UI) that allows them to work line by line in a way that looks like they are texting.
So the key innovation in the translation industry is going to be a new UI for translators? We’re going to upend the world by moving from a document-based approach, or a table-based translation approach, to a chat-based approach? … What if they want to go back and fix something? What if they want to review a few paragraphs ahead in order to gain some context? Let me go out on a limb and say that a chat-based interface, whatever its positives for other usages, is not a good translation interface whatsoever. Even if it were, in and of itself it is a minor element of innovation in the translation business.
First, Stepes designers had Myers’ questions already in mind when designing the app. To review texts for context, Stepes translators simply swipe left and right on the phone to reveal the entire original source or translated text. To edit an existing translation, the translator simply has to press on the text bubble. By tapping into common existing mobile behaviors, Stepes delivers better user experience even when compared to traditional desktop translation software, leading to improved translation efficiency and quality.
Secondly and most importantly, the real innovation behind Stepes goes beyond its chat-based interface; it’s how Stepes will enable new models of translation and thus change the way we all think about globalization and language. Specifically, Stepes and other mobile-based technologies will make translation bigger than ever, one that scales up to a level that actually matches global communication needs. By making translation more mainstream, we will usher in an era of “Big Translation.”
The world certainly isn’t lacking in language talent. Nearly half the world speaks two or more languages – that’s 3.65 billion people alive today with the potential to contribute to translation. Unfortunately, a number of factors have conspired to make translation an unappealing career choice. Among them are: cumbersome and overly technical translation software platforms, and the perception that the industry is not exciting enough. Very few children reply “translator” when asked what they want to be when they grow up.
Fortunately, there are a number of innovations that are trying to make translation faster, more affordable, and more accessible. Chief among them is the groundbreaking idea that we can bypass traditional translation tools and move to a mobile app platform. Stepes is easy enough to use that anyone bilingual and with a smartphone can become a translator and earn money. With the right tools, translation doesn’t have to be a demanding full time career but rather an activity that everyone engages in at some point in their day.
If anyone bilingual could translate straight from their mobile phones while, say, waiting in the grocery store line or on the subway, then we could pool our language talent that otherwise would have gone untapped and vastly increase our global translation capacities. In redefining what translation looks like, we can drive down global communication costs and raise translation’s public image. This is the democratization of who can translate and who can afford translations.
In the era of Big Translation, businesses could post projects of any length and in any language that would be picked up and completed by translators working around the world, that very day. Big Translation would combine aspects of machine translation with the sensitivity of human translation. We could finally make translation fast and inexpensive but also guarantee the complete accuracy of human translation.
Excitingly, we would also be able to build a global community around translation. With a Big Translation approach, translation would be a collective, global effort. We could have millions of part-time translators, located across the world yet simultaneously working among the word’s language pairs. When more people translate, we get better results. Subject matter experts – doctors, engineers, artists, students – could translate content that they are passionate about to achieve higher accuracy with content containing industry vernacular. Progress would also accelerate; as we translate more content in more languages, we can also begin to build up a global database of standard translations for specific terms and phrases, speeding up future translation efforts.
By harnessing the collective language talents of people around the world, we will be able to drive down translation costs for businesses while also empowering bilingual people to translate in their free time and earn money. In this way, Stepes users can “share” and pool language abilities that would otherwise have gone underutilized to tackle large translation projects for globalizing businesses. Importantly, Stepes doesn’t treat translation as a commodity – we realize that translation is full of nuance and context. What the app is allowing people to share are their language abilities in ways never possible before. Meanwhile, translations are still subject to the same quality control mechanisms as before.
AirBnB and Uber have already successfully tried this sharing economy model with travel accommodations and taxi rides. Could it work for translation? The answer is yes, because the translation process can be even more easily completed on a mobile end-to-end without requiring a car or a real estate property. All you need is your smartphone, a device most adults already own. In this sense, uber translation is more uber than Uber.