Translation as a profession seems to be on the minds of many, and it's reflected in the many books that offer career advice to translators. In the past few months I've reviewed these three books by active translators who many of us know well:

• Alex Eames' Business Success for Freelance Translators: How to Build and Run Your Own Freelance Translation Business
• Judy and Dagmar Jenner's The Entrepreneurial Linguist: The Business-School Approach to Freelance Translation
• Chris Durban's The Prosperous Translator: Advice from Fire Ant & Worker Bee

Now it's time for the last of the publications that I'd been planning to review, this one written by Corinne McKay and titled  How to Succeed as a Freelance Translator (2nd edition). Corinne is a veteran writer about translation, known to readers through her excellent blog and her now-defunct newsletter on open-source translation tools that I often referred to (she stopped writing the newsletter when she discovered blogging).

As with the other books I've reviewed, this is a pleasant read. If I were a translation newbie, I'd be plain stupid not to invest in this book. The checklist in the all-new Chapter 3 alone ("Your First Year as a Freelance Translator") would make it worthwhile. Corinne includes many more helpful tips, as well: how to write a résumé (though I wish she'd stuck with the first edition's spelling of the word rather than this edition's resumé); advice on contacting translation agencies; and suggestions for building relationships with direct clients (my favorite section). From this last tip you can see that this book is not only for novices -- experienced translators will also profit from its advice. It's good for all of us to be reminded of (or taught) good marketing practices, including the use of social networking (I loved the rather extensive section on LinkedIn and the clever advice on Facebook). It's also helpful to receive advice on seemingly mundane things like obtaining health insurance or raising your rates. (On second thought, why exactly did I call these things "mundane"?) Another very useful tool is the table on page 147 that will help you discover the hourly rate you need to charge -- you might be surprised to see that you've been significantly undercharging.

The book is primarily geared toward the translator living and working in the US, but Corinne has tried to include others as well -- though she could do only so much in 200 pages. I must admit I was a little disappointed with Corinne's section on tools. Though she often used the term "translation environment tool," from my admittedly biased perspective she focused too much on the tools' translation memory features when other features, such as terminology maintenance and QA, are just as important. I think this is important to understand, especially for newcomers.

With that one criticism aside, I really liked the book. In fact, I'm already looking forward to the next book that she plans to release later this year with a summary of her blog postings.

I'll leave you with a nugget of wisdom that I thought was not only very clever but was also an aha moment for me. On page 26, Corinne writes: "Localization breeds localization; a localized web browser automatically creates a need for localized websites; a localized piece of software demands a localized manual to go with it."

If that's not a happy thought for a language professional, I don't know what is.

SDL Trados Studio 2009 Freelance

If you have been waiting to invest in Studio 2009 then now is the time to take action! 

Until 31st May we are offering you an amazing 40% off new licenses and 30% off upgrades, make sure you don't miss out!

 To buy or upgrade visit our dedicated portal -- www.sdl.com/promo/toolkit5  

This is an admittedly silly title because encrypting files and folders on your computer -- especially on a laptop that you schlepp around with you all the time through unsecured wireless networks -- is not such a bad thing to do, especially when it comes to data of a more confidential nature.

Whether it makes you paranoid or simply wise, here is how to protect those files without the help of a third-party product in Windows (7 Professional/Ultimate, Vista Business/Ultimate or XP Professional) and without any kind of fuss.

To encrypt a file or a folder, simply right-click on that particular file or folder, select Properties, click the Advanced button, select the Encrypt contents to secure data checkbox, and then click OK.

That was easy.

If there is a lot of data in that file or folder it may take some time for the encryption process to complete, but you'll know it's finished when the encrypted folders and files appear in a different color than your other files. And that's really all you will notice -- opening, saving, or even emailing them will all work as before (they will simply lose their protection once you attach them to an email). Don't believe that they're protected? Try to log on with a different user account and you won't be able to open them (that's also why it's not possible to encrypt commonly shared files such as system files).

Now if you don't want to click through various dialog boxes to encrypt or decrypt your files, you can also add a shortcut menu item to Windows Explorer.

To do this you will have to open the registry (select the WinKey -- the one with the Windows icon -- or open the Windows Start menu), type regedit and press Enter.

Now navigate in the Registry Editor to:

HKEY_CURRENT_USER/Software/Microsoft/Windows/CurrentVersion/Explorer/Advanced

Right-click on Advanced and select New> DWORD (32-bit) Value from the shortcut menu. On the right side of the window you will now see New Value #1. Overwrite that name with EncryptionContextMenu and press Enter. Once that is done, double-click on this value and enter 1 in the Value data field. If you now click OK and leave the Registry Editor, you will see a new Encrypt command in Windows Explorer when you right-click on any file or folder icon -- and if the file or folder is already encrypted, you'll see Decrypt.

Be safe!

As some of you are now slowly moving toward Trados Studio, here are a few resources for those with a practical turn of mind.

Many of you know Tuomas Kostiainen, one of the stateside Trados gurus -- in fact, maybe the US-based Trados guru. Tuomas is a ubiquitous presenter at ATA conferences and at meetings of the many local chapters of the ATA, including the Colorado Translators Association earlier this week.

Well, that's helpful, you might say, though the information's a bit dated. True, what's done is done, but there are at least two interesting resources that are still accessible. One is the conscientious Twitter stream of Eve Bodeux, who religiously noted all the cooler tips and tricks that Tuomas shared. The other is a copy of the presentation by Riccardo Schiaffino and Margherita de Togni about some of the flaws of Trados 2007 in comparison with Trados Studio.

And to come back to Tuomas (or as they would say in Finnish style: Andtocomebacktotuomas), he also has a very helpful blog that details some of the struggles and questions encountered when switching from Trados 2007 to Trados Studio. Much recommended!

I know, I know, some of you might have marveled at the cool Twitter query for filtering Eve's postings from the CTA conference that's listed above (and, yes, I also know that most of you didn't give a hoot about it). Anyway, Twitter has some search operators that we all know from other searches -- such as quotation marks for an exact match of a phrase, a minus sign for excluding a word, or OR for searching for one or the other of two terms -- but there are some that are shockingly different than most others we're familiar with. Here's a list of those (each of these is to be entered in the search box atop Twitter's main window):

• from:jeromobot -- sent from person "jeromobot" (cool account!)
• to:natalykelly -- sent to person "natalykelly"
• "good food" near:"chicago" -- containing "good food" and was sent near Chicago
• near:NYC within:15mi -- sent within 15 miles of "NYC"
• ATA since:2011-05-13 -- containing ATA and sent since "2011-05-13"
• ATA until:2011-05-14 -- containing Jeromobot and sent up to "2011-05-14"
• movie -scary :) -- containing "movie," but not "scary," and with a positive attitude
• flight :( -- containing "flight" and with a negative attitude
• traffic ? -- containing "traffic" and asking a question
• "Tool Kit" filter:links -- containing "Tool Kit" and linking to URLs
• translation source:twitterfeed -- containing "translation" and entered via TwitterFeed

Good luck filtering!

I started reading your book last night and I LOVE it. It has all the information I have always wanted to know. I will definitely be recommending it left and right.

Excellent job!!

Charlotte Brasler (English<->Danish)

The Translator's Tool Box: A Computer Primer for Translators

400 pages with (almost) everything you ever wanted to know about your computer  

In the last newsletter I wrote about tools that can be used to change many file names at one time. Reader Amal Ibrahim asked about a file renaming utility that allows him to append the date to file names.

One of the tools that I mentioned -- 1-4a Rename -- allows for exactly that but manages to hide it quite well.

After you start the tool (there is no need to install it -- simply download it and extract it somewhere on your computer), you'll need to press F2 to switch into "Expert mode." Once you've done this, select the location of the files you'll need to change and apply appropriate filters (such as only .txt files or only files that start with "a") and then right-click on the Insert/append field. In the ensuing submenu, select Filename> Date> Date, make sure to select the correct position in the file underneath, and click Start.

Your changes will be applied in the blink of an eye. At that point you may notice that there is no undo and you should have done a better job selecting the right kind of files. . . . So, be careful!

ErrorSpy is one of a handful of tools that quality check the files coming from your translation environment tool (TEnT), and it did a fine job of convincing most TEnT developers that they had better improve their own quality assurance features. Until a few years ago, this feature had been offered only by stand-alone tools such as ErrorSpy, QA Distiller, and Xbench, but now most major TEnTs include many of those QA features by default. Naturally, this is bad for the sales of those stand-alone tools, but it's good for us because we didn't have to purchase yet another tool.

However, you won't be surprised to hear that the makers of ErrorSpy and QA Distiller in particular have not sat still; in fact, they've enlarged the repertoire of the checks they offer quite significantly (I've written about that before).

Just last week, ErrorSpy was also released as an online tool: ErrorSpy Online. That's good because a) it's free (if you don't want to customize the settings, you do only one file or file pair at a time, and you don't upload very large files) and b) it will give you a good chance to see the tool in action.

The tool allows you to upload your bilingual files (in the formats .ttx, .xlf, .sdlxliff, .csv, .txt, or .tmx) or a translated file pair (in the formats .txt, .doc, .docx, or .pdf), optionally a glossary in .csv, or .txt format to check adherence to the glossary terminology in the translated file, and then send them to be processed. Within just a minute or less you get a very comprehensive report that contains information about:

•  missing or unmatched 
• incorrect number format
• inconsistent acronyms
• empty or identical target segment
• target segments that are too long or too short in comparison with the source
• file-internal consistency of the translation
• plus, if you have uploaded a glossary, a report on terminology adherence.

Unfortunately, the supported languages are German, English, Spanish, and French only at this point, but if you do work in those language combinations it might be worth a try.

In the last newsletter I mentioned that I had talked to Systran, the market-leading rules-based machine translation tool vendor. Interestingly, I received a good number of emails just for mentioning Systran in my newsletter, mostly from readers who commented on Systran's lack of understanding for the common workflow of translators and fairly inadequate training. I forwarded some of the messages to Systran in the hope that these issues will be addressed. I would understand if Systran truly does not understand the common translation workflow -- but now, as they discover that their tool has an increasing attraction to some translators, that would certainly need to be changed.

One thing that I was particularly interested in during our conversations was their claim to be a "truly hybrid system." This would typically mean that all language processing is done with a combination of rules-based machine translation (RbMT) and statistical machine translation (SMT). "Hybrid" is the favored term and concept right now since that is what most developers are looking at.

Note that even Andreas Zollmann of Google said earlier this year

that the idea that more and more data can be introduced to make the system better and better is probably a false premise [that's the concept of statistical machine translation -- Jost]. "Each doubling of the amount of translated data input led to about a 0.5% improvement in the quality of the output," he suggests, but the doublings are not infinite. "We are now at this limit where there isn't that much more data in the world that we can use," he admits. "So now it is much more important again to add on different approaches and rules-based models."

This was a rather ground-breaking statement from Google, which thus far had firmly banked on SMT.

To come back to Systran's hybrid system, their answer was that their desktop products -- while using a customization "which learns from documents and emails (...) and analyzes documents located in specified folders and creates dictionaries and languages models" as well as TMs from aligned documents -- don't really use SMT. It is only the enterprise server that "includes techniques and tools for SMT training." I'm glad we got that clarified because I was pretty certain that a normal, non-dedicated desktop computer would not have the processing power to handle the very large amounts of data that are needed for SMT, and I was confused by the claims about it being a "truly hybrid" system.

In contrast to PROMT, the product that I talked about in my last newsletter, with Systran the SMT results "are used for statistical post-editing of RbMT results, in addition to source adaptation and for the creation of syntactic language models." In a way, PROMT does it the exact opposite way by having the SMT results first and using the RbMT results to refine those. (All of this is a little over-simplified, but this is just for illustration's sake.)

In response to the translation memory feature that Systran offers, the folks from Systran pointed out that they use fuzzy matching (but only down to 97% accuracy) as well as some sub-segment matches, both, to my knowledge, features that are not used by PROMT.

I also asked them about the underlying engine for Babelfish which is powered by Systran and other online offerings and the use of statistical machine translation there. And it looks like this is not going to happen -- in fact, their plans for the future are "highly specialized translations for each domain (...). The ultimate generic SMT is not our objective." But "we focus our efforts on  deeper integration with machine learning techniques that are part of SMT."

I thought all of this was quite helpful.

As far as the supported languages go, Systran supports English, Chinese, Dutch, French, German, Greek, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, and Swedish in various combinations; PROMT supports English, Spanish, German, French, Portuguese, Russian, Italian, and Lithuanian (and Chinese in the server editions) in various combinations as well. (It's very important to evaluate the different language combinations independently -- the results can differ significantly.)

As a subscriber to the Premium version of this newsletter you have access to an archive of Premium newsletters going back to May 2008.

You can access the archive right here. This month the user name is toolkit and the password is privilege.  

New user names and passwords will be announced in future newsletters.

If you would like to promote this newsletter by placing a link on your website, I will in turn mention your website in a future edition of the Tool Kit. 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.

© 2011 International Writers' Group