You can view earlier editions of the Tool Box Journal going all the way back the 2007
in the archives to which you have access if you support my work on the Journal.

Tool Box Logo

 A computer journal for translation professionals

Issue 18-12-294
(the two hundred ninety fourth edition)  


1. Guest Article on Pricing for Translation by Michael Schubert

2. An Artificial Intelligence Proposal

3. The Tech-Savvy Interpreter: The Webaround -- Green Screens and Chroma Key Technology for the Masses

4. Vice Precognition

5. Disable Fuzzy Repetition

6. This 'n' That

The Last Word on the Tool Box

Wondrous Things

Here's a story that at first glance might not make sense in a technical newsletter; but then maybe that's exactly where it should be. At the ATA conference in New Orleans, something happened that still makes my heart go pitter-patter when I think about it. About six months ago somebody asked me whether I could think of a way for him to donate a bunch of Russian dictionaries that he no longer needed after retiring from his translating life. After going back and forth a bit with him and the ATA conference organizers, we decided to host a dictionary exchange -- or, more accurately, a dictionary giveaway since there was no need to bring something in order to take something. I don't think I'm sharing any confidential information by revealing that there were great doubts, and maybe even concerns, that this would completely flop. Truth be told, I wasn't quite sure whether it would work either. Those old paper-based things? Who still uses those?

Well, for me, and I think to some others as well, it was really one of the highlights of the conference this year. Over the three-and-a-half days of the conference, people brought between 300 and 350 dictionaries, including stunning specimens such as a Bulgarian-Albanian dictionary, formerly highly confidential Russian military dictionaries, and other beauties that had been languishing in the bookshelves of their former owners. All but eight of the dictionaries were taken (I took those last few to a nearby used bookstore and donated them), and it was so much fun to see the joy on the faces of those (of us) who just could not believe their good fortune at being able to take as many treasures as they wanted.

The Bulgarian-Albanian dictionary may not see much use (but then, who knows!), but I know that many other handsome tomes will be used again -- much to the joy of their new owners, their generous previous owners, and probably the dictionaries themselves.

Here is a view of some of the excitement after the first batch of dictionaries had arrived:

dictionary exchange 

Enjoy this season and this (very long) Tool Box Journal that includes a guest article by Michael Schubert. Be well.


Webinar: Successful socializing for localization professionals

As we enter the season for celebration and socializing join this webinar with Judy Jenner for tips on making the most of social marketing and entertaining as a way to grow your network and make new connections in 2019.

Sign up here >>


1. Guest Article on Pricing for Translation by Michael Schubert

In one of the many great discussions initiated by a curator of TranslationTalk, this one about rates, Michael Schubert chimed in the other day with this:

I began phasing out word prices in 2010 and stopped using them altogether in 2013 -- even with agencies. It began with my refusal to engage in a conversation around "Trados rates" or haggle over pennies. Shifting the focus from word count to content changes the whole conversation!

I thought that was interesting, so I asked Michael to write something up for the Tool Box Journal, and he was kind enough to oblige. Here it is:

Changing the Conversation on Pricing

What's in a word? Precious little. We all know that we don't translate words, or even sentences and paragraphs: We translate content, in all its context. (Worte and not Wörter, for those of you who know German.) And yet, most of us bill our services by the word. Why? And what message does this send?

The why is simple: It's the easiest way to assess the volume of the task and provide a price in advance. And the message? To our clients, it says we are selling a commodity -- the very message that, elsewhere, we keep trying to contradict. To ourselves, the message is that the faster we can translate, the more money we will earn. It incentivizes speed, not quality.

Moving away from word rates toward hourly or project-based fees better reflects how we actually work -- and how we want our work to be perceived.


When I entered the profession, I worked mostly for language service providers (LSPs) and charged by the word. As I moved upmarket, this model became unsustainable for me: Since my word rate had risen to become higher than most, LSPs saved the tricky or sensitive jobs for me, leaving me feeling underpaid. Or they peppered me with smaller jobs that, even with a minimum fee, led to death by a thousand cuts. And then came the dreaded conversation about scaled word rates based on TM analyses (more on that below), which is what really drove me away from that model. Word rates also discourage, or make it difficult to bill for, value-added services: documenting errors or ambiguities in the source text, consulting with the client about the objective and target audience of the text, suggesting adaptations ("transcreation"), and so forth.

I began phasing out word prices in favor of hourly and project-based fees in 2010 and have not used word rates at all for many years now. LSPs generally ask for a word rate, since this is the dominant model and allows them to more easily compare translators by price. My response is always that I cannot offer a one-size-fits-all rate for future jobs sight unseen, and this argument resonates with LSPs that truly care about quality and have this same conversation with their end clients. Hourly and project-based pricing changes how clients think about our services, underscores that a 500-word internal company memo is not the same as 500 words of website content, and opens the door to offering and billing for value-added services, which helps us move upmarket.

Direct clients

As we approach and begin working with direct clients, we need to think carefully about how we present ourselves. Are we like the paper clip vendor who deals in unit prices, discounts by volume, and works with some low-level vendor manager? Or are we more like the consultant who provides professional services and interacts with department heads and senior management?

Word pricing mischaracterizes the nature of our work and leaves us talking about pennies. And that defines us as vendors of a commodity with a low unit price.

The businesses I work for don't think in terms of word count, anyway. My clients never say, "Can you translate 2,000 words by Thursday?" They say, "We have a white paper we'd like you to translate." And from there, I ask them about target audience and deadline, not budget. Only after I've seen the text and talked with them about the project do I offer a price quote. By then, I've amply demonstrated that I care about their baby. This is how we change the conversation from unit pricing to value.


I would (and do) argue vehemently that translation environment tools do not, on balance, make us faster -- they yield smarter and more sustainable workflows, improve internal consistency, and give us more versatility and efficiency in working with complex file formats. Proficiency with state-of-the-art translation tools should never be a trap door to lower prices. It is, in fact, a rationale for higher prices. (Please reread this paragraph three times. I'll wait.)

So instead of trying to hide my use of TEnT tools from my clients for fear of opening up the dreaded conversation on scaled word rates, I make sure my clients know that I use state-of-the-art translation technology to organize their bilingual content and harmonize their corporate style over a relationship that will last years, even decades.

Last word

We know that translation is not a commodity, and we need to demonstrate that in how we frame the conversation around our services with clients. Word count may be the metric of choice in the language services sector, but we translators are the largest group in this sector and have it in our power to steer the dialog away from individual words and toward content -- away from quantity and toward quality.


Michael Schubert is an ATA-certified German-to-English translator based in San Francisco providing premium translation services with a focus on corporate communications, information technology, and finance.



🏝💻  Celebrate 20 Years of Translating Faster with Wordfast 🐠☀

Join Wordfast for their 5th annual user conference in Martinique on March 21-23, 2019 

👉 Nearly 20 sessions and workshops to learn how to use Wordfast better

👉 1-to-1 meetings with developers and experts 

👉 Snorkeling/scuba diving, rum distillery visit, beach volleyball, networking dinners, and more...

Register today: 


2. An Artificial Intelligence Proposal

The other day I received an email from a colleague lamenting the complexity of one of the leading desktop-based translation environment tools. She was especially exasperated by its intricacy after having primarily worked in web-based tools that tend to have a much simpler interface.

These were some of her main concerns (and these are actual quotes):

  • The cluttered interface that literally hurts her eyes
  • The multitude of features that do not relate to the translation process
  • The tool tip boxes that randomly appear and obscure her view of the text she is trying to type

She asked several times in the email whether she's the only one who feels that way. She's not. In fact, I would dare to say that especially now, more and more translators feel the same when faced with essentially two classes of software products: the all-bells-and-whistles, desktop-based (with some cloud features), "traditional" tools; and the cloud-based tools with a more minimalistic interface.

It's probably important to ask where the increasing complexity of the first class of tools originated. I think there are primarily two reasons.

The most obvious, of course, is our diversity and our large variety of needs. Not only do we translate in many different language combinations and subject matters, but our positions within the translation process and corresponding needs are wildly different, including anything from translator and editor to project manager to vendor manager or IT and database administrator. All of these groups need to be satisfied with their specific set of features.

The other reason is something very positive at its core: competition. Developers have naturally attempted to best one another with an ever-increasing set of features -- individually helpful and valuable features, at least for one segment of the user group. But with each additional feature, the tools have continued to balloon to the unwieldy size we have today. Developers are not unaware of these problems. Both Trados and memoQ, for instance, have focused specifically on usability in their latest releases.

For years I've been defending the difficult learning curve and multi-screen approach of existing tools by saying, "If translation were an easy task, the technology that supports it could be easy also. But since we know the former is not true, how can we expect the latter?" And while there is still a grain of truth there, it has seemed increasingly hollow. Yes, the technology that drives the tools supporting the translation process is by definition complex, but does that mean the user interface has to be complex as well? I don't think so.

In talking with software developers, I have often suggested that they offer several kinds of user profiles for their users to choose. Star Transit has taken this approach, though not in quite the way I've imagined as most helpful. They focus on functional roles (terminologist, translator, reviser, etc.) with the necessary features presented once the role is chosen. But why not (also) use the same concept for technological savvy? Clearly, any role the user chooses would be reversible, and any feature the tool offers could be activated at any point.

This seems like a no-brainer to me, and I find it kind of hard to understand why no one has ever taken that leap.

However, I recently realized that the availability of new technology might achieve us even better results with something slightly more advanced. Yes, I'm talking about artificial intelligence (AI).

Let me very quickly interject a note on the kind of artificial intelligence I'm talking about. AI experts talk about different kinds of artificial intelligence: "narrow AI" and "general AI." Narrow AI is when a technology makes a prediction based on previously processed data within a very specific field (think AutoComplete or Netflix suggestions). Computers have been doing this for a long time, and it works really well. General AI is when a computer becomes aware of its environment and makes appropriate decisions accordingly -- comparable to what humans do. As translators, our experience with the unpredictability of generic machine translation has proven that general AI is still far from employable. (A recommended read on the subject of artificial intelligence and the typically male-dominated world of "technochauvinism" is "Artificial Unintelligence: How Computers Misunderstand the World" by Meredith Broussard.)

Returning to translation environment tools, tools like Memsource and Lilt have prided themselves for implementing artificial intelligence features unrelated to the "usual suspects" like neural machine translation. Instead, they've incorporated it into other features such as language-specific recognition of untranslatables or morphology engines. In translation project management there is also a lot of talk about using artificial intelligence to streamline and automate workflows and ease the job of project managers.

Here is my challenge to tool vendors who are selling technology that causes users to respond like my exasperated colleague: Why not allow a user's actual interactions with the system to determine what features the system should offer to that user? If the user never works with translation packages because she works only with direct clients who don't prepare translation projects, there is no reason to have the whole variety of options dealing with processing packages available to that user. If the user never uses the termbase (shame on him!), why not make the user interface less cluttered by hiding all the relevant features? I could clearly go on and on.

Naturally, I'm not the first to come up with this idea (for instance, see this article), but I do think this would be a good approach, especially in a field like ours. A highly complex tool like Microsoft Word offers the user an overwhelmingly large number of options. In contrast to a translation environment tool, however, the processes activated through Word's many features are successive, one after the other, on a single screen. We expect our translation environments to process many things simultaneously because we want that multitude of resources displayed as we translate. But we don't all use the same number or kinds of resources. Similarly, when processing XML files, it's admirable that even the most advanced users will find all the options they desire, but why not have a simple layer below all this to make it accessible for the less-technical user as well?

In the recently published ITI Research Network e-book, two of the authors discuss the sometimes problematic collection of data during the translation process. I agree, it can be potentially problematic if it's done for surveillance purposes. But it's important to remember that the kind of data collection required for an AI-driven customization of our translation environment user interface would be completely different. In fact, it will be just like the existing "Usage Data Collection" in memoQ or the "SDL Trados Studio Experience Program." Both of these optional programs transmit data about the use of their tool back to the developers so they can gain insights on what features of the tool are used and in what frequency. What if that kind of data could be used to customize the tools to the individual process and preferences of each translator?

Of course, this could work just as well for browser-based tools; in fact, there it would be even easier for the developer to collect such data.

Users -- in this case, translators -- need to be included in the development process. Part of that process may be using AI to take an active role in designing our individual tools to our individual needs.



Across Online Training for Translators

Are you new to Across, and would you like to get help to get started with the Across Translator Edition? In our four-hour online training, our trainer Julia will walk you through the software. Learn more 


3. The Tech-Savvy Interpreter: The Webaround -- Green Screens and Chroma Key Technology for the Masses (Column by Barry Slaughter Olsen)

The growth of videoconferencing and video remote interpreting (VRI) have laid bare an inconvenient truth -- most of us don't have a professional-looking place in our homes where we can participate in web meetings or interpret remotely. I grappled with the problem when I first started teaching interpreting online in 2016. My home office is a little overpopulated with my books, my printer, my kids' trophies, family get the point. A cluttered background is distracting and just doesn't look professional. To address this issue with their interpreters, especially sign language interpreters, who work from home, VRI companies usually require them to have a screen or backdrop behind them.

There are many ways to get a simple monochrome backdrop behind you for VRI work, but most of the solutions that I explored initially were too cumbersome or required a permanent change to my office. I settled on a pop-up green screen and stand that are about 8 feet high and 6 feet wide when in place, but they take up a lot of space in my office and require some setup and takedown time. And when up, there isn't much room in my office for anything else.

Another cool thing about these screens and backdrops is that you can use them together with chroma key technology to insert virtual backgrounds of any color and style into your outgoing video. It's what most television meteorologists use when they provide the local forecast each evening. Once available only to motion picture and television studios, chroma key technology is now within reach of anyone with a modern computer, webcam and green screen. But few people want to paint their home office walls bright green or hang a large green cloth on the wall behind them. I've been using chroma key technology for a couple of years now with web conferencing platforms like and the open-source streaming and recording software OBS. It provides a much more professional look and keeps the home office clutter out of view.    

I was content with my large green screen until, while attending Zoomtopia in San Jose, California, in October, I came across an ingenious backdrop solution for VRI interpreters who work from home called The Webaround. It's a foldable backdrop that mounts on the back of just about any office chair providing you with an instant uncluttered backdrop for video remote interpreting. (If you having a hard time visualizing just what I'm describing, go to their website and check it out. A picture is worth a thousand words.)

So, as soon as I returned from the conference, I purchased one to test it. I've been using it regularly for the past month or so. Here are some observations:

  • The Webaround is lightweight and easy to set up and take down. It takes me less than 30 seconds to open and install it on the back of my chair. It folds and unfolds like those sunshades that you put in your car windshield on a hot, sunny day to keep the dashboard from getting too hot. That makes setup and takedown of the backdrop easy.
  • It fits snugly on the back of my office chair. The model I purchased (the 56-inch Big Shot) comes with an integrated stabilizer to keep the backdrop from drooping so it stays correctly positioned behind you.
  • Even though I bought the largest chair-mounted backdrop they make, I still had to zoom in my Logitech C920 webcam to get the green screen to fill my entire video background. The need to tighten the shot to get full coverage means that these chair-mounted backdrops may not work well for sign language interpreters who need to have as wide a shot as possible, so their clients can see what they are signing. But for spoken-language VRI, the backdrops seem ideal.
  • The green version (they also come in blue and grey) of the Webaround works well with chroma key software, so I can easily project different backdrops behind me, depending on which client I am meeting with.

All told, I really like my Webaround. It has quickly become my go-to green screen for impromptu web conferences, online teaching and remote interpreting, mainly because of the ease of setup and use.

Do you have a question about a specific technology? Or would you like to learn more about a specific interpreting platform, interpreter console or supporting technology? Send us an email at



Registration for Lenguas 2019 Now Open. Register Today!

We're just two months away from this one-of-a-kind conference. The program is published, and workshops are filling fast! Continuing education credits for ATA and CCHI have been approved. CIMCE credits are in process. Choose from dozens of continuing education workshops for translators and interpreters. Join your colleagues in Mexico City on January 24-26, 2019!

Click here for more information.


4. Vice Precognition

Whether I wanted to or not, with my left hand temporarily out of commission, I've had to once again solely rely on voice recognition for the last couple of weeks. If you don't find many errors in this Tool Box Journal, it's due not to my (non-existing) above-average skills of dictation but rather the above-average skills of my editor. The truth is that voice recognition is simultaneously an incredibly powerful tool and frustratingly confusing. Not confusing because it's difficult to use -- it's anything but -- but because of the different strategies of proofreading. There are no squiggly red lines below typos -- the program writes out only known and correctly spelled words -- and while it might highlight syntactical and grammatical errors, these checks are not available in many programs and are often not particularly helpful anyway.

When I mentioned on Twitter the other day how welcome a webinar on proofreading dictated text would be, a good number of colleagues chimed in with suggestions, including the tried-and-true method of having the text read back to you by one of the many tools that can do that. And, yes, that can be helpful, but it doesn't catch homophones, which are one of the biggest problems when it comes to voice recognition. Here's what I have found helpful: changing the background color of the text I've dictated (in Microsoft Word, for instance, you can do that without changing anything permanently by selecting View> Read Mode> View> Page Color> Inverse). Somehow this tricks my brain from reading the text as something I entered where errors are magically hidden into something that can be evaluated more objectively. Most importantly, though, I've learned to distrust my texts more and read them more slowly, a fair trade-off considering that I've gained a lot of time by dictating rather than typing.

I've also noticed that I'm noticeably better in German dictation than in my non-native English. It's that little bit of an accent and my German inability to distinguish between aspirated and non-aspirated closing consonants in English that penalize me with the additional couple percent of errors.

Of course, I'm among the fortunate who deal with two languages that are excellently supported by Dragon, but the good news is that there are more and more possibilities for many other languages as well. I've previously mentioned GBoard, Google's keyboard for mobile devices, for its huge number of languages that can be connected to transfer to your computer (see the 290th Tool Box Journal), and just last week Kevin Lossner mentioned on his blog the voice to text notebook as another possibility.

But even if you try to control your computer as much as you can with voice recognition, there's always the random keyboard command that you want to enter directly on your keyboard. And if you have a virtually unusable hand like me, those are the times when you start to hate keys like the CapsLock or Insert key that your clumsy hand might activate all too often.

Finally I found the simplest but very handy software to deactivate or reassign those keys. Believe it or not, it's called CapsLock Goodbye and does exactly what the name describes. (Tip: if you use it, make sure that you remember where you downloaded it or stored it on your computer since you might need it to reactivate those keys at some point.)




PDF Translation for Professionals

Translating PDFs is easier and quicker with TransPDF.

  • Available within Memsource and memoQ
  • Compatible with all CAT tools.
  • Fast log-in for Proz members.

FREE for your next PDF project.  

All file formats, all languages, all target groups, better quality, shorter time-to-market -- one standard solution.


5. Disable Fuzzy Repetition

Kevrenn DFR is a little program -- or really an application-driven process -- that is interesting partly because it grew out of an actual need perceived by the translation agency Kevrenn.

They were frustrated by the translation projects they sent to freelance translators that typically contained a large number of segments they called "fuzzy repetitions." These are segments that are different only because of the difference in numbers or punctuation or a single term (provided that the term is either classifiable as a variable -- such as a <product name> -- or is contained in the termbase). While different translation environments are able to locate those "fuzzy repetitions," Kevrenn's program goes one step further by preprocessing any translatable file and locking each fuzzy repetition so translators see the context but are not able to do anything to the translation of those "reps." It's naturally not paid for either. Once the translated file or project is sent back to the agency and the translation memory is populated with the translation data, either the project manager (if she is familiar enough with the languages in question) or an editor fills in the empty and now unlocked rows.

I had a relatively long talk with Jean-Marie of Kevrenn about his system, partly because it took him a while to convince me of the benefits (and I'm still not quite sure he succeeded). I can see the system working well for smaller operations where there is a real or perceived need to squeeze cost as much as possible without immediately interfering with the pay to translators. On the other hand, I imagine that quite a few translators would find an option like this not particularly helpful to the workflow and would not be thrilled to work with it. I think this is one of those situations where much depends on the relationship of the translator and the agency they work for and goodwill on both sides.

The tool is not immediately for sale, but Jean-Marie would be happy to talk with you about how you can use it either as an SaaS solution or as a desktop-based tool that so far has been tied in to memoQ, the translation environment that Kevrenn uses and for which the solution was originally developed.



Your style, your wording, your content. Utilize translations quickly and reliably throughout your company.  

Corporate machine translation powered by STAR MT

Watch the short video for more information on STAR MT functionality and usage:


6. This 'n' That

You'd have to have been on NASA's InSight on its long and finally completed journey to Mars to not have heard that IATE, the massive European Union terminology database, has been completely overhauled -- at least as far as the user interface is concerned, including the possibility of connecting to it via application programming interface and an overall greater usability. Kudos to the team that worked on it: I think the result looks beautiful. Over the years I've had the opportunity to meet some of the people who are responsible for developing and maintaining IATE, and I have rarely encountered such passionate linguists.

Some of you who don't use IATE directly via its interface but through tools like IntelliWebSearch might have been a little frustrated when you realized that the old links don't work anymore. Both the folks at IATE and IntelliWebSearch (and I assume any of the comparable products) are working on a fix for that which should be available in no time.


GT4T -- the little tool that connects you from any program to many machine translation engines, glossaries, and other online resources -- is now available to Mac users as well (and by the way, already supports the new IATE).


I'm looking forward to a free webinar I'm giving to launch a series of translation community-oriented webinars sponsored by Lilt. I will be talking about the Translation Technology Wiki, what the platform can do, and what can be done to make it even better.


With the holidays upon us and so many colleagues to thank and show appreciation to (remember that large job last April you never would have finished on time if he or she hadn't helped you?), how about a surprise gift for them? Or something for yourself?

Here's a one-of-a-kind gift idea: Aside from his regular presence on Twitter, Jeromobot -- the robotic, eight-centimeter tall, wind-up patron saint of translators -- has gone very quiet. Well, it turns out there was a reason for his silence: He's been multiplying, and now he has 30 tiny clones! Every single one wants to find a new home, and each comes with a CD or USB stick containing the Translator's Tool Box e-book (value $50) plus a signed copy of Translation Matters (value $15) (signed by me, not Jeromobot!). Until December 15 (or until the 30 translator-extraordinaire packs are gone) they're available for $30 (plus postage). Click here to order or find out whether there are any packages left.



The Last Word on the Tool Box Journal

If you would like to promote this journal by placing a link on your website, I will in turn mention your website in a future edition of the Tool Box Journal. Just paste the code you find here into the HTML code of your webpage, and the little icon that is displayed on that page with a link to my website will be displayed.

If you are subscribed to this journal with more than one email address, it would be great if you could unsubscribe redundant addresses through the links Constant Contact offers below.

Should you be interested in reprinting one of the articles in this journal for promotional purposes, please contact me for information about pricing.

© 2018 International Writers' Group    

Home || Subscribe to the Tool Box Journal