A while back when I was explaining our industry to a friend, she was surprised at the complexity of our work and quipped that I should write a book called Translation for Dummies. I replied that there are no dummies in the language industry, only an astonishingly diverse community of highly-qualified and super-motivated people. When I went on to mention that there are close to 7,000 languages spread across the globe, she was floored, and stammered in amazement that it’s a miracle we actually manage to get anything done.
It was all light-hearted, but it got me thinking about the sophistication of what we actually do and how can we explain it. The truth is, there are no shortcuts. Best practices are more like rules of thumb. And whatever our methods actually are, there is often more than a touch of madness in getting projects out the door.
Since the tech we now use is very well-structured, it might seem like a good idea to follow that in describing how we use it. But are the processes of translation best captured in this way? Not really. Whatever claims are made for machine translation (MT), the business of making natural language conform to the hard logic of computing is still problematic and always will be to some extent. This is forcefully illustrated by the skepticism the language community expresses when claims of parity with human translators are made.
Language quality can never be dismissed simply as a given; it has to be crafted, maybe even obsessed over. I learned this firsthand when I applied for ISO 9001 certification in medical translation for my language service business back in the 1990s. We labored to describe in meticulous detail every stage our translators and ancillary staff had to go through in delivering projects that fully satisfied our clients’ rigorous requirements — not to mention the International Organization for Standardization (ISO) standard of guaranteed perfection every time. The resulting manual inspired our IT people to speculate about using it as a basis for automation, but the tech back then was in its infancy and we had neither the time nor the resources to even think about it realistically.
While strong leadership skills are a given for the C-suite, one of the pitfalls of existing in the rarefied atmosphere of the top business echelon is that the technical skills they used to climb up the corporate ladder can be left to their subordinates. Given that there are only 24 hours in a day, this may be excusable. But if those in leadership roles want to issue decrees about rolling products and services out globally, they must keep their fingers on the pulse of what is feasible. Throwing cash at internationalized language services may not bring the results they intended. Remember Frederick Brooks? He is known for declaring: “adding manpower to a late software project makes it later.” Well, the mythical universal language translator, I would argue, fits just as well.
For instance, business goals are often set by marketing, one of whose functions is to identify opportunities. However, to identify India as a massive potential source of revenue without addressing the multilingual needs of its 22 official languages is a painfully inadequate generalization, and charging ahead without addressing this is not something that could be fixed in a last-minute scramble.
Internationalization has been recognized as a major force that can be harnessed in directing business and cultural ambitions and most enterprises have their own teams dedicated to this. But in conversation with many diverse professionals, I am still not convinced that we’ve encapsulated all that internationalization implies. For example, what similarities do different internationalization departments share? All the colors of the business endeavor rainbow are here, so what processes do they have in common? It’s perfectly natural that they tailor their activities to their specific businesses. But the myriad different languages and the cultures that use them do not exactly conform to the same classification.
In truth, for some people, grasping what is involved in a basic translation workflow and what happens in a linguist’s mind would be progress.
The technology now available is not monolithic. It covers many functions and for our purposes may be regarded as working in two distinct activities. First, we have what are referred to as productivity tools. These, of course, require some interaction with users. The benefits are well understood and all translation professionals use them as a matter of course. The other is MT.
Project management, we tend to forget, used to be an awful lot simpler. But requirements used to be a lot simpler. Now workflow has superseded project management with its deliverables, milestones and life cycles. Software engineering underwent the painful process of learning that projects are not beasts that are happily caged and easily controlled. Integrating the entire supply chain into one workflow is the holy grail of translation, with some commercial applications doing it better than others. But this involves everyone working on the same platform, which presents its own problems, security being at the top of the list.
It is this technological capability that excites corporate executives and causes jitters among language professionals. This is what has forced many in our community to cast their future as merely post-editors, and their gloom is well-founded. But even tech A-listers like Jaron Lanier, the virtual reality pioneer, are indignant that the human sources of the material co-opted into translation memories and so on are disregarded and go uncompensated for their work.
We used to use phrases like “quick and dirty” or “gisting” to describe work that put deadlines or project costs ahead of quality. It certainly is true that quality is not always the prime aim of a piece of work; sometimes clients simply need a good idea of what a source text means. That acceptance of a less-than-perfect standard gets around difficulties such as style in some scientific and technical writing, but such compromise depends upon the end-reader understanding and accepting that this is the case. There are, of course, many occasions when audiences will not accept substandard quality. Think of all those internet posts featuring unintentionally comic translations. Getting it right is important, and in my experience most if not all professional translators strive to do just that.
If the language community expects satisfying transformations in our working lives to emerge from dictates issued from the top floor of corporate headquarters, we might be disappointed. We as a community must at least contribute to our own destiny, if not forge it anew altogether. If we wait for emerging technologies to show us the way, we will be missing a golden opportunity to embed the enabling benefits of multilingualism across the board. There are no dummies in translation. We are highly-qualified professionals who provide the widest range of services in the world. It’s time to translate that across the globe.
Incubators in our industry
While the yin of theory lets us probe all kinds of wonderful ideas, we need the yang of practice to really test if they work, so let me turn to a real-world case in point.
The Process Innovation Challenge (PIC) was first staged at LocWorld34 Barcelona in 2017. Disruptive innovation seems to have originated in the pages of the Harvard Business Review in the 1990s, but the principles apply just as much to the localization industry as anywhere else. Dave Ruane, a veteran of our community and member of the Enterprise Business Team at Xplanation Language Services, devised the event after being inspired by a similar effort in another industry. With the backing of Ulrich Henes, the event was staged, immediately hailed as a success and is now an established feature of all LocWorld conferences.
So, what exactly is PIC? Like TV’s Shark Tank, it is a competitive event that gives localization innovators the opportunity to pitch their ideas, pitting them against other rival hopefuls. As Ruane explained to me, “There are nuggets of innovation in every part of the industry, some of which could be easily used or repurposed by others, they just need to be seen.” A chosen panel of “process dragons” fire expert questions at contestants. The winner is selected by audience vote. What better way to source ideas ready for incubation than to showcase them in front of important members of our community?
But incubation involves a lot more than simply providing a hatchery, then counting chickens. PIC is now in its second year and the goal is to grow and evolve the platform. At present, PIC is single-track. However, with an abundance of diverse processes and the spread of technical tools at our disposal, the possibility of multiple tracks is being investigated. Growth seems assured. As Ruane told me, “Anyone can innovate, the trick is to share innovative ideas in an applicable and usable form that someone can repurpose or use as a starting point for their own ideas.” This is not about randomly sowing seeds, this is to invoke Henry David Thoreau’s disruptive ambition of making “the earth say beans instead of grass.” In today’s language community, we have a plethora of talent to stage our own innovation moonshot. “Some really great ideas get spotted at the PIC, someone invests and in turn this idea drives the next level of disruption in our industry,” as Ruane puts it. While he is not predicting what this will be, he believes it is just a matter of when.
This is not merely wishful thinking. In my work with the volunteer community of Translation Commons over the past few years, many issues have surfaced with real-world requirements that need to be dealt with through new processes housed in a commercial setting. Over the past year, the advisory board has started Parallelle, an incubator to generate solutions for our industry. Just as Translation Commons was created in response to actual community needs, Parallelle is creating new workflows, processes and frameworks that address issues for which current business practices have proven ineffective.
Responding to the frustration of translation buyers who struggle with problems not of their own making, I have to emphasize how critical it is simply to listen. Diagnosing problems cannot proceed otherwise. But prescribing treatment for these problems sometimes requires some original thinking, some pioneering chutzpah to deliver the goods!
We all know the temptation to resist change because “we’ve always done things this way,” or “don’t fix what ain’t broke!” But in our fast-paced, multifaceted, globalized world, we need to be attuned to what drives the clamor for new and enhanced products and services. The stronger our collaborative efforts, the better our chances of evolving linguistic applications that satisfy global demands.