Interview with Jost Zetzsche: “I am a TRANSLATOR and I say it with great pride!”
Last month we had a chance to spend an hour with Jost Zetzsche and ask him a few questions regarding his own experience in translation industry, the future of translation, CAT-tools, machine translation, and a few other things. You can listen to this interview online. We have also put the concise text version below. We think the interview turned really insightful and useful for both translators and translation buyers. Here it is, enjoy!
Olga Arakelyan: Hello everybody! This is Olga at Alba Longa Translation Company, and we are pleased to tell you that we decided to have a series of interviews with some very important people who play an important part in the translation industry. And our first guest today is Jost Zetzsche. Jost, hi! Thank you so much! You woke up so early to talk to us!
Jost Zetzsche: Well, thank you for having me. I appreciate it.
Olga: For those who don’t know Jost yet, he is an ATA-accredited English to German translator, a consultant in the field of localization and translation, and a writer. Jost writes on technical solutions for translators. Jost is an author of two books, and both are very popular. One is “A Translator’s Toolbox: A Computer Primer For Translators” and the second one is “Found in Translation”. Jost, from that book I have an impression that you are completely fascinated with translation and localization. How did that fascination start? How did you get into the industry?
Jost: Well, my background is actually in Chinese. I studied Chinology, or Chinese studies, in Germany and spent some time in China. And then at a certain point I focused on the Chinese translation as a field of study, so I wrote my PHD theses on the history of the Chinese Bible translation, which in some way has very little to do with what we do today, because what we do in the translation and localization industry is very short-lived, while the Chinese Bible translation took a couple years to be finished. So, from that sense it’s the opposite kind, but it taught me a lot of translation principles and it gave me a certain grounding as far as what translation quality is and how to approach translation in general. After I finished my PHD I moved to America and I first wanted to continue with the academic path, but then I started to work for a translation company and then after a couple of years made myself independent and I’ve been independent ever since.
Olga: Wow, that’s amazing. But now you are more engaged in all the different technical solutions. As you know, some translators treat new solutions and software as their worst enemies, while others embrace them. So what’s your opinion on the role of the technology in the translation and localization process?
Jost: It does play an important role. In general we do accept technology now as something that we have to use and many of us also welcome it as something that’s a positive thing to use. The reason why I started writing about technology was to help translators to accept technology more and to become sort of an evangelist for technology. And I still am in a way, I think, but I’ve also come to realize that what we call industry is in the reality a very wide field with all kinds of different needs and there is not one technological solution for everyone. I know plenty of translators who make a very good living in what you and I would consider to be a very low tech kind of way. And there’s a place for those people as well. I think that if you work in certain areas where you work with certain kinds of texts, you are extremely well-advised to use technology and most translators know that. But there are other areas where you can go by with a very minute kind of technology. So yes, technology is important, but to assume that technology is equally important for everyone or that everyone needs a certain set of technology – I’ve sort of come away from that concept.
Olga: I see what you mean. I read quite a few blogs and journals that translators write and I see that very often colleagues who are more or less well established (and maybe older in age), don’t like technology that much and they prefer not to use it, while the younger generation is generally more friendly towards it.
Jost: Right. I think what you also see on blogs is that the translators who are more experienced and who are maybe into technology often assume that their particular set of technology is the right set of technology for translation and they sort of have this missionary zeal to say, “I found the solution and this is what everybody should use”. I am oversimplifying here, but there is that mindset sometimes and it’s not really helpful. Technology is moving along. We are now looking at a completely different set of technology that will have an impact on some of us. We need to be able to not look at what we’ve been able to achieve in the past with technology. It’s by definition something that is developing into new areas, and so we need to be brave and open enough to look into new areas as well I think.
Olga: And that’s what you are doing with your Toolbox Journal, right?
Jost: I’m trying. I’m just as limited as anyone else is, but I am trying to be as open as I can for new developments and I am trying to find out what the benefits or the drawbacks of those new developments might be not for me personally, but for translators in general. I am trying to talk to the developers of the various tools and technologies very early on, long before they actually release the technology, and to help them to steer the technology in the right direction. Well, into the direction that I think is right. It’s sort of the role that has developed out of writing about translation technology, and and I think if it is helpful for the developers and hopefully helpful to the community as a whole.
Olga: Well, I think it is very helpful. As far as I know you are the go-to person concerning many technical issues. What I have noticed also from your biography is that you have seen the translation process from different perspectives. You’ve been a translator; you’ve been a project manager. Now you are supplying the translation industry with the software solutions. So you have seen this whole process with the eyes of at least three different people. Here’s a tricky question: which of those roles do you like best and why? And another question would be: How different are those views actually? How differently do those people see the translation process?
Jost: I would always prefer to be a translator. What I liked about being a project manager was working with many, many different languages. And I still do that today as I help companies when they have technical difficulties. But the reason why I prefer to be a translator is that I love to work with language. It may be frustrating to translate the same kind of manuals over and over again, we all know that, but still I love to work with language and that’s the reason why when somebody asks me what I do, I always say “I am a translator”. And I say it with great pride.
Another area that I feel very strongly about is the famous answer “I am just a translator”. I feel very strongly that it’s a very inappropriate answer. I am not just a translator, I am a TRANSLATOR and I am proud of it. I think young translators don’t realize it’s really important to be a translator. No matter what kind of stuff you translate, you are playing a very important role and I happen to be really proud of that role and that’s why I can without reservations say that I’m proud to be a translator. I don’t always enjoy it, there’s always stuff that’s boring. But overall, I really enjoy translating and of course I try to do other things as well, such as writing about translation and technical solutions. That’s something I would always encourage translators to do: to translate and then maybe also look for something that’s related to their work as translators, but gives them a thing that they can do when there are not as many jobs coming in, or if they really feel they have to do something else after the monotonous translation work on a manual or whatever they do!
Typically we have certain specialties as translators. We specialize in certain fields, whatever that might be. So it makes a lot of sense to take those specialties and develop other fields of expertise aside from translation in those areas and have that as maybe the field where you have a second leg to stand on. And I am not really saying that from the economical point of view, but from the peace of mind and diversification point of view. I think that’s something that I would really encourage translators to do.
Olga: I see. Well, I have another question for you. What do you think about machine translation? Especially the English-Russian language pair? What do you think is the future of machine translation in this language pair?
Jost: I don’t speak and read Russian, but as you know I’ve been to a number of conferences in Russia and Ukraine in the last few years and so I’ve spoken to a lot of people who have looked into English-Russian and Russian-English. I think that it’s very similar to German, which also is not particularly well-suited for machine translation in comparison to some other languages. Languages that are morphologically rich, like Russian, or German, or others, those languages are much more difficult to translate with machine translation or to have an accurate or an almost accurate machine translation.
I think that we have made a mistake to look at the machine translation mostly from the post-editing point of view. So we look at the machine translation as something when we get a suggestion from machine translation and as professional translators or post-editors we then are to correct, or to post-edit, that translation. I think in some instances that might be the right way of using machine translation, especially if you have extremely well-trained machine translation engines for a specific purpose or a specific sort of text. But in other cases that might not be the best way of using machine translation.
But even in those cases there might be ways of using machine translation appropriately by weaving it into your translation process. Machine translation can come as a productive new element that helps you in your translation process. Some tools are now starting to use interactive kind of machine translation where you receive suggestions from machine translation engines as you translate, that you can either completely ignore, or you can accept. So you don’t necessarily focus on a complete sentence or segment, but maybe just a short sub-segment that your machine translation set might actually have right. And you can accept just those 3 or 4, or 5, or 6 words. Or if you have processes by which you have matches from the translation memory, which is a database of your previous translations, and those matches might be fuzzy matches (matches with slight difference to the original sentence). And the machine translation comes and corrects those because it might know that term or that phrase that is missing or incorrect. So those are things that I think have a lot of potential for us.
Like many of us, I’ve worked as a machine translation post-editor and, to be honest, I didn’t very much enjoy it. Some people really enjoy it, and I think that’s great. I didn’t happen to enjoy it very much maybe because the engine I was working with was not particularly good, maybe because my language combination isn’t as well-suited as others, but I am very excited about some opportunities that machine translation brings us. I think we need to look at creative ways of using machine translation and to integrate it into our process. That’s the difference – we sometimes look at machine translation as machine translation’s process where we play a little role, but I would like to see it as our process, the translation process that is guided by the translator who has just been given different tools to work with. And machine translation could potentially be one of those tools that could help you.
Olga: Wow, that’s an amazing picture that you just drew. I can also see where machine translation could be useful. So far it’s very far from being ideal of course, and even from being good in some cases and some language pairs. It would really be good if we could use this blend of translation memory and machine translation, plus our own input of course, because nothing can really substitute a human being in this process.
Jost: Right, and I actually think that for those who are really interested in the financial bottom line, if we find really clever ways of doing that we would be more efficient, than if we would be post-editing not particularly good machine translation output. Again, if you have extremely well-trained translation engines, for those situations it might indeed allow translators to post-edit it and in many cases you don’t need much post-editing for simple sentences etc., but if you have more complex texts, if you have a language combination that might not be as well-suited or if you don’t have those specifically well-trained engines, which most of us as translators don’t have access to, then I think we will find much more productive and satisfying solutions by looking at those combinations.
Olga: Now let’s talk about CAT-tools. First of all, you have written a lot about different CAT-tools. Which program do you personally like best and you prefer to use in your translations? And which would you recommend for startup translators?
Jost: As for tools, I use a great variety of CAT-tools. In fact, I often look for jobs with CAT-tools that I haven’t used in the productive setting just to get an idea of how those CAT-tools work in a truly productive environment. Traditionally I have used Déjà Vu because I have a long relationship with both Emilio and Daniel Benito who stand behind Déjà Vu and I’ve written all the documentation for them, so I am very familiar with the program. So I often use Déjà vu, but truth is I also use a lot of other tools.
What I don’t typically recommend is an easier tool for beginners. I think beginners should choose the tool that helps them. It’s ok to change tools in your translation career or to have several tools, which I think is the case for many of us. But I would not say, “Well, let’s start with XYZ and then later on you can go to ABC.” I would say that beginners should choose the tool that they could also imagine to be using 10 or 15 years from now. And I think if you choose a tool, what you need to do is look at your environment. You need to look at the clients you are working for. Do they ask you do use certain tools? And if so, is it possible to still use other tools? You know sometimes clients would say, “I want you to use Wordfast or Trados, or MemoQ”. But is it possible to still use a third-party tool that is also able to convert the Wordfast or Trados, or MemoQ files into something else?
And the next question would be “What do people use that you work closely with?” You want support. Not just the official support that the SDL, or Kilgray, or Wordfast provide, but you also want support from your peers. So if your peers all use a certain tool, that’s a pretty good indication you should use that tool as well because they have a good experience with it, and also because they are the ones who’s going to give you support. And only then, after you answer those two questions, you should make a decision for a tool I think.
Some of the tools are quite amazing. I think tools like MemoQ, OmegaT, Trados or Déjà Vu, and some of the new web based tools like Wordbee, XTM or Memsource are amazingly powerful and are to different degrees constantly evolving. If you look at MemoQ, for instance, they are constantly adding new features. If you look at Trados, they are constantly evolving by having third-party developers develop solutions for them that take care of certain aspects of processing and of translation. OmegaT is an open source program that has new releases every other week. So there are a lot of different tools that are worthwhile to look at. And as you know, if you are well-situated and you have work for some time, you end up having several tools on your desktop anyway. You won’t just have one tool, but you will have a number of tools to choose from. Some of them might be supplied by the clients, some might be free to start with, and some of them you might have to purchase.
Olga: Yes, I know that. Did you have a chance to work with the first Russian CAT-tool ? It’s called SmartCat.
Jost: I haven’t worked with it, but I had a number of talks with ABBYY about it. And I think it’s really interesting. SmartCat probably has the potential to become the tool that people will be talking about quite a bit. I think the reason why it’s particularly interesting is because of its pricing strategy: it’s free for translators. And because they are able to use some of ABBYY’s core strength like OCR, Optical Character Recognition, and integrate it right into their CAT-tool. So they are able to easily process image files and probably use the best conversion solution for PDF files. And that’s just part of their CAT-tool, so it’s really interesting. I think that one thing that will not be particularly easy for them is that it’s a web based tool. Like any other web based tool vendor, there’s still a certain hump to take for companies to use cloud-based solutions. We are not particularly progressive as an industry when it comes to accepting solutions like that. The cloud-based concept has not really taken hold I think. But it will, so it might be just the right time for them to launch it now and to do quite well.
Olga: Why do you think are the cloud-based tools not so popular?
Jost: Well, one of the hesitations that translators have, which is understandable, is that we work for a large variety of clients and many of our clients are very concerned about privacy. Many of our clients have very strict guidelines of where texts can be located. Some of our clients have very strict regulations that none of the texts can be sitting in company external servers, which excludes the clouds. So that’s a good reason for being skeptical about cloud-based solutions.
The other reason is that kind of a nebulous skepticism about cloud-based technologies in general, and that I don’t quite support as much. I don’t understand why we have to be so nebulous about it. We all use Gmail in some way or the other, or Yahoo mail, or Hotmail, and that’s also cloud-based, and the attachments to the clients end up on those servers also. We need to be more clear on what our problems with it are and why we have those problems. Again, some of those problems are understandable and I see where those concerns come from. Others are not as acceptable as concerns.
Olga: Yes, I see what you mean. I also thought about this privacy issue which is of primary importance to all translators in our work.
Jost: Of course, it is. We need to take our clients very, very seriously. And if the client says “no’, then it’s “no” in this particular case. There are other things where we can argue with the clients, but not on those matters. But there will be a change with those clients who now say “you may not store anything on external servers. A couple years ago many of clients who now say “it’s ok” used to say “It’s not possible with our documents”. Obviously the whole world is moving into the cloud, and so will our clients more and more I think.
Olga: Well, speaking about CAT-tools, what is very interesting for me is the reason why Trados is so popular. A lot of translators, translation companies, and large corporate clients ask for Trados, they prefer to work in Trados or to have others work in Trados for them, so what is the reason for that?
Jost: Well, I think there are both historical and technological reasons. Historically, Trados had a quasi-monopoly on translation technology. That has changed; it’s not a monopoly any more. SDL is still the market leader and partly for good reason. They have technology that’s quite powerful and caters to the different groups within our industry. If you oversimplify it, there’s a translation buyer, and then there’s an LSP and then there are translators. And all those three groups find their needs met to some degree in SDL translation technology, including Trados, Word Server, and the other SDL solutions.
I think those are two main reasons why Trados is still quite dominant. Besides, if you have a production chain like we do, it takes a lot of effort to change the technology that the production chain is using, because if one part changes it, all other parts have to change it as well. Overall, I think that since Trados Studio appeared in 2009 there have been a good number of changes, and the monopoly of Trados is broken. That’s a good thing. It’s good for Trados and it’s good for other tools as well. Monopolies are not helpful for anyone. I have nothing against Trados. Like I said, it’s a technologically very good solution, but I do have something against the market where one technology really dominates and has the ability to essentially dictate what’s happening in the industry. I think that’s not the case anymore.
Olga: And I know that you even initiated a campaign gathering ideas for CAT developers. Is that right?
Jost: Last January I did a couple of webinars where we collected ideas about things translators want to see in CAT-tools that they don’t see right now. So we collected a good number of things and I sent those on to the CAT-tool developers and asked for their response to that. They sent responses and quite a few actually replied to specific suggestions saying “Yes, it was a good suggestion and we are going to work on that, and you are going to see that within 6 months or a year in our tool”. Others were not quite so positive. And some said, “Well, we actually have that feature and nobody knows about it”. That was also good, so now we know about it. And just this morning actually I talked to the person who organized those webinars and we are going to have a webinar again next January where we are going to revisit what has become of those promises and where we are, and what new has happened in our industry and technology that would require maybe new ideas that could be brought to the developers.
Interestingly, one big area of concern for the participants in the webinar was voice recognition. The ability to use voice recognition within CAT-tools in not particularly strong. So many translators wanted to see a much stronger integration of voice recognition within CAT-tools. That was one big area.
Another big area was of course machine translation. And the general user-friendliness of tools. If you or anyone who’s listening to this wants to see all the different suggestions and the responses of all the CAT-tool vendors, they can just send me an email. I’ve compiled them all into one big excel spreadsheet, so I’d be happy to send it.
Olga: Oh, that would be great! It’s very interesting to see what people think, because it also speaks a lot about the basic development in the industry.
Jost: Yes, I think that this effort also represented an opportunity to the developers to look beyond their immediate constituencies. Their immediate constituencies are the existing customers. And they have certain requirements or wishes, and tool developers might or might not respond to those wishes. But this was an opportunity to look at all the potential customers who are not using it right now and who do not talk to those developers. They suddenly had a voice, they wanted to be heard and those requirements or wishes were communicated to developers and they all responded very positively. I think that we also have not helped ourselves sometimes to not particularly trust the developers. We should really look to work with the developers to take the technology to what it is and we’ll find that developers are just as open to it as we are.
Olga: I see. In the today’s translation market we also hear more and more about those programs that are called translation management systems, like XTRF for example. They help both companies and individual freelancers to automate the translation process. Do you think it makes sense to use these solutions, especially for freelancers? Do you have your own personal favorites among such programs?
Jost: Well, I know that both XTRF and Plunet are thinking of ways to market them for freelance translators. I don’t see an immediate need for those tools for freelance translators. There’s sort of a line between freelance translators and LSPs is not particularly well drawn. Some translators also forward work to colleagues or to their peers. Once they do that, I guess they become a LSP. If you are a strict freelancer who only works on translation and doesn’t share with colleagues, I am not sure that you need a program like that. Plunet or XTRF are very powerful for language service providers who have a great throughput of jobs and need ways to handle those jobs in a more uniform and more streamline manner. But for freelancers – I don’t really think it’s important. For freelancers, I think it’s important to have a really good organization of the jobs they are receiving and working on. They need to have a good overview of the billing things and they need to have structure as to what kind of prices for what kind of clients etc., but you don’t need a large ERP solution like Plunet and XTRF.
Olga: Now let’s speak about conferences. We know that you take part in many international translation conferences. There is a wide variety of them now. But sometimes it feels like these conferences now duplicate one another. So is it all about networking or is there really nothing new to say about translation business? Which conferences do you find more useful or more interesting?
Jost: I think you are right when you say that conferences are about networking. They are about networking. And I am talking from the translators’ perspective now. It’s important for freelance translators to go to conferences to build up strong networks of peers that will support them in their work as translator and will possibly help them build virtual teams of translators for large jobs etc. So that’s a great value of conferences. I think you are absolutely right that there are ever more conferences and if you look at the speakers of those conferences, very often you see the very similar kind of speakers if not the very same kind of speakers and they often don’t say particularly new things. But of course that’s an overgeneralization and you do hear interesting things at conferences.
That’s probably not something that can be generalized for everyone, but what I thought particularly interesting is to go to conferences in countries or areas where there have not been many conferences before. For instance, it was thrilling to be at the Translation Forum Russia a couple of times, or at the Ukrainian conference a couple years ago, because those were conferences in the areas where nothing like that had ever happened before. I am going to a conference in Czech Republic in September. I’m looking forward to it, because again there’s great excitement, and that excitement that participants have also reflects on the speakers. And maybe new things will come up at conferences like that. That doesn’t mean that the FIT conference in Berlin just a couple weeks ago wasn’t a fantastic conference. Lots of really important things were said there, and lots of great contacts were made. The ATA conferences that I typically go to (this year it’s going to be in Chicago) are also very important conferences. So I think you should have a big set of conferences to go to. Typically you cannot go to 6 or 7 conferences a year because it’s expensive. So if I were to advise somebody, as far as conferences are concerned, I would choose one large conference like the ATA or whatever is in your area that you live in, and one conference of your regional translation group, like the Translation Forum Russia, or the Ukrainian conference, or that kind of new and exciting conferences.
There are also the Localization World and GALA conferences, but those really are more geared more towards language service providers and they have a different kind of importance for those folks.
Olga: I see what you mean. Well, speaking about translation industry and just to sum up our interview, it’s clear to everybody that translation and localization industry is going through some very drastic changes now. Where do you see the industry in 10 years from now? Or if 10 years is too much, maybe 5 years from now? And where do you see yourself?
Jost: I think for the industry, there are really interesting changes ahead. We are going toward more diversification right now. I think that 10 years ago a LSP could say “I am offering translation into too many languages”. That might have been good enough, but I think it’s not good enough anymore today because the concept of translation in the public’s mind has changed quite a bit.
What do you hear about in the media about translation? You hear about Google Translate, about these tools like Skype with the new built-in interpretation. You hear about those things, so as language service providers we have to become much better at explaining what we do. And by doing that I think we’ll find out that we are actually doing very different things from each other. So we have to actually find out who we are individually. Not as a whole, but individually. Talking about stories, I think we have to find our stories. And I think here is a huge opportunity of providing very specialized services for very specialized clients.
So we have to find who we are. And what that will result in will be diversification. Those who can’t define who they are shouldn’t be here in the first place, because if you don’t know who you are, if you don’t know what services you really want to offer and what you are good at, then what are you doing here to start with? The majority of us sort of knows who we are, but we haven’t actually put it into words, and that’s what we need to do. And if we do that right, we are going to see a very rich landscape that we used to call ‘the industry’. And we will find out that there’s actually a great variety of services that we offer. And that’s why I am struggling with the term “industry”. Industry sort of presupposes that there’s a monolithic kind of landscape and it’s not! It’s a very, very rich and diverse landscape of services and things we can offer.
When people ask me about the future they also ask me about machine translation. You know, I think machine translation is going to play a role, but not a role that has to worry us by any means. What has happened recently is that an American company called Smartling has got a lot of outside investment – 60 million dollars. And I think they will play a big role, for good or for worse. They are likely to challenge SDL’s role as a technology provider.
But overall, we are going towards the future that’s quite bright, but not the future where it’s enough to say “I am providing translation”. It’s a future where it’s important to say “I am providing a service that I would like to explain to you. Here’s what it is.” And then you explain it in an understandable way. Translation industry is about communication. That’s what we do – we communicate between languages. But terms like “localization” or all the other terms we’ve been using are not good terms. We are well-advised to use language that clients understand. Our clients are really smart, but if we come up with new acronyms and new terms all the time that are just internal to our own particular world, that’s not going to help our clients to understand who we are. We need to talk in an intelligent language that our clients can understand. I think that’s really important and we haven’t done that very well in the past. We will, and things will get better if we do.
Olga: Jost, thank you very much for taking your time to answer all my questions. The interview was insightful and very informative. Thank you very much.
Jost: You are very welcome. It was fun talking to you.