CHRONICLE - Finnish Translation Services and Trados Training

October 2006 Volume XXXV

Number 10

A Publication of the

American Translators Association

In this issue:2006 Conference CoverageNisei Linguists During WWII Questions: An Editor’s Best FriendInterpreter Certification Programs in the U.S.

CHRONICLETheJanuary 2007

Volume XXXVI Number 1

A Publication of the

American Translators Association

ATAJan07-bi-REV2.qxd 12/27/06 9:48 AM Page 1

The ATA Chronicle ■ January 20076

Our Authors January 2007

Nataly Kelly is the director of productdevelopment with NetworkOmniMultilingual Communications. She is also acertified court interpreter (State ofMissouri) for English and Spanish. Aformer Fulbright scholar in sociolinguistics,her current research interests are inter-preter certification, quality improvement

programs, and telephone interpreting. She currently serves on theNational Council on Interpreting in Health Care’s Outreach Committee.Contact: [email protected].

Tuomas Kostiainen is an ATA-certified(English→Finnish) translator specializingin technical and medical translations. Inaddition to freelance translation work, healso provides TRADOS training and consul-tation, and has given several presentationsand workshops on the subject. He is thecurrent president of the Northern

California Translators Association. Contact: [email protected].

Rina Ne’eman is a Hebrew translatorand interpreter with 25 years of experi-ence. She is the managing director of RinaNe’eman Hebrew Language Services, Inc.She has translated and interpreted forprime ministers, major television networks,Fortune 500 companies, and premier lawfirms. She is a member of ATA’s Public

Relations Committee. Contact: [email protected].

Jacki Noh has been a freelance Koreantranslator/interpreter/voice-over talent forover 20 years. She is a member of theInternational Association of ConferenceInterpreters and a former member of theJudicial Council of California CourtInterpreters Advisory Panel. She has alsointerpreted during the post-medal press

conferences for three Olympic games (Atlanta, Sydney, and Salt LakeCity). She is currently serving on ATA’s Board of Directors. Contact:[email protected].

Kayoko Takeda is a freelance English↔Japanese interpreter andtranslator. She teaches interpretation and translation in the GraduateSchool of Translation and Interpretation at the Monterey Institute ofInternational Studies (MIIS). She holds an M.A. in translation and inter-pretation from MIIS, and is currently enrolled in the Ph.D. program intranslation and intercultural studies at the Universitat Rovira i Virgili inSpain. She is working on her dissertation on the sociopolitical aspects ofinterpreting at the Tokyo War Crimes Tribunal. Contact:[email protected].

It’s Time To Renew

If you have not renewed your membership, now is the time!

From membership in any or all of ATA’s 15 divisions—including the new Language Technology and Korean

Language Divisions—to discounted business services,ATA gives you the strategic edge that benefits

your bottom line.

Renew online at:www.atanet.org/membership/renew.php

Thank you for your past support and for renewing for 2007.

Continuing Education PointsVisit www.atanet.org (click on certification)

Guide to ATA


23The ATA Chronicle ■ January 2007

Using TRADOS is likepicking berries in a Finnish forest. Ifyou know the forest and where to findthe best berry patches, you will do fineand get a lot of berries in no timewithout getting lost. However, if you donot know your way around or even whatto look for, you will end up withoutberries, frustrated, lost, or perhaps evenworse, harassed by a bear. Rather thanrisk this plight, many choose to stayclose to the forest’s edge, making dowith the berries they find there.

While teaching TRADOS work-shops and providing one-on-onetraining, I have noticed that manytranslators tend to stay within theperiphery of the first useful functionsthey encounter, unsure of how to ven-ture into what they perceive as a forestof options to get to the more advancedfeatures TRADOS has to offer (thevery best berry patches). “Oh, I don’tknow where to go,” or “I can’t readthe map,” they say. I realize that as atrainer, my job is not to teach every-thing that TRADOS does, but to tellworkshop participants about the bestberry patches and draw a useful mapso that everyone can get there andback, directly and safely. So, here ismy collection of some of the best fea-tures TRADOS has to offer, alongwith a map for finding them.

Memory or Memories?Many (if not most) translators use

several different translation memories(TMs), creating a separate memory foreach client or subject. While theremight be some good reasons for thisapproach in certain situations, I find it

much more efficient to use one largememory with attribute fields (such as“client” and “subject”). This allowsyou to utilize all of the previous trans-lations you have done for other clients,

or those covering a particular subject. Having a large number of client- or

subject-specific memories defeats, tosome extent, one of the main purposesof using a TM tool like TRADOS.Some programs, such as Déjà Vu,allow you to use several memories atthe same time, but TRADOS does not(except for concordance searches,which I will explain later). Combiningeverything into one large memorydoes not have to be as messy as itsounds, just as long as you classifyyour translation units (TUs) accordingto client, subject matter, projectnumber, and/or any other attribute ortext field you might find useful.

Using Attribute Fields Attribute and text fields (together

called TM fields) can be used as selec-tion criteria when searching a TMduring a translation or when per-forming memory maintenance. How,you ask? Just bear with me while I takeyou a little deeper into the forest—itshould all make sense in a minute.

For example, during translation,

the “Attribute and Text FieldDifferences Penalty” option inTRADOS (Options > TranslationMemory Options > Penalties) givespriority to TUs that have the same

attribute/text field values as your cur-rent project by “penalizing” (= low-ering) the fuzzy match levels of allother TUs by two percentage pointsper field (this is the default value, soyou do not need to do anything to setthis up).

For utilizing attribute/text fields inconcordance searches, you can selectthe “Apply Current Filter Settings”option (Options > TranslationMemory Options > Concordance) ifyou want to give a priority to TUs thathave specific attribute/text fieldvalues. And as I mentioned earlier,these fields can also be used whileperforming memory maintenance asfilter criteria when you want to utilizefind and replace operations or export acertain subsegment of your TM, forexample (File > Maintenance >Filter…).

Here is an example of how to setup attribute fields and their values(see Figure 1):

1. In TRADOS Workbench, select“File > Setup… > Fields.”

➡

Having a large number of client- or subject-specificmemories defeats, to some extent, one of the main

purposes of using a TM tool like TRADOS.

PickingBerries With TRADOSBy Tuomas Kostiainen



2. Select any field in the “AttributeFields” list.

3. Click “Add” and type a new name(such as “Subject”) into the “Field”input box. You now have a fieldcalled “Subject.”

4. Select “Subject” in the “AttributeFields” list and then “A Value”from the “Attribute Values” list.

5. Click “Add” and type a new name(such as “Medical”) into the“Field” input box. You now havean attribute field called “Subject”containing one value called“Medical.” Repeat this for addi-tional fields and their values.

6. When you have added all fields andtheir values, click “OK” (do notclick “Add” at this point anymore).

Using Text FieldsText fields function differently, in

the sense that you set only the textfields but not their values in advance,and then write the needed value in the“Project and Filter Settings” dialogbox (see Figure 2). An example of atext field could be a project number oranother piece of information that isused only once. A project number isused only for that project, but a sub-ject name (= attribute field) will beused every time you translate for thatsubject.

If you have not yet used attributeand text fields, it is worth spending afew minutes to figure out what type offield structure would be most useful inyour case. Two or three attribute fieldscould be enough, such as subject andclient. Of course, you can add newfields and their values later.

Selecting the Appropriate FieldsThe only downside to using

attribute/text fields is that you have toremember to select the appropriatefields and their values before eachnew project. The selection is done asfollows (see Figure 2):

1. In TRADOS Workbench, select“Settings > Project and FilterSettings.”

2. Select appropriate attribute/textfields and their values for projectsettings and filter settings. Note thedifference between the two:

• Project settings define the fieldsand their values that will beadded into each new TU.

• Filter settings define the fieldsand their values that will be usedas filters when searching the TMduring translation.

3. You can get the “Project and FilterSettings” dialog box to appearautomatically every time you openTRADOS or a new memory byselecting “Options > TranslationMemory Options > General >Show Project Settings” (you needto do this only once).

4. Remember to change your projectand filter settings if you start workingon another translation project withoutclosing the TM or TRADOS.

5. Also, remember to verify your cur-rent project and filter settings beforedoing an analysis or clean up.

Sounds like a hassle? Yes, a little,but it pays off quickly. It is like usingmosquito repellent in the woods—ahassle, yes, but you will be happy you did.

Figure 1: Fields tab open for adding text fields and attribute fields and their values.

Picking Berries With TRADOS Continued



Segmentation RulesTRADOS uses so-called segmen-

tation rules to segment the source text.By default, segmentation rules arebased on punctuation marks so that asegment is normally equivalent to onesentence. This works fine in most sit-uations, and one rarely needs tobother with these rules. However,there are situations where adjustingthem slightly can make your workmuch more efficient.

Using Stop Characters: I have trans-lated numerous ingredient lists for var-ious products (pet food, nutritionalsupplements, etc.) where each listoften includes pretty much the sameingredients, but in a different order. Ifyou translate these lists using defaultsegmentation rules, each list is treatedas one translation unit, and you end uptranslating each ingredient again andagain without benefitting from yourprevious translation. However, if youinclude a comma or semicolon(depending on which one is used toseparate the ingredients on the list) asa “stop character,” you can translatethese lists ingredient by ingredient. Asa result, by the time you start workingon the second ingredient list, you willget many 100% matches. You will alsonot have to figure out how to translate“partially hydrolyzed organic Balticherring liver oil” or retype it again (seeFigure 3).

Changing Segmentation Settings: Tochange the settings, go to “File >Setup… > Segmentation Rules.” Youcan either create a new category foryour stop character (such as a commaor semicolon) or just add it under“Marks” together with the questionmark and exclamation point that arealready there. The “SegmentationRules” dialog box might look a bit con-fusing, but do not be too afraid to playwith it. With the Reset “panic button”you can easily reset the default settings,and no harm has been done.

Overriding Segmentation Settings:One easy but often overlooked featureyou can use during translation to over-ride the segmentation rules is the“Expand Segment” (Ctrl+Alt+Pagedown) or “Shrink Segment”(Ctrl+Alt+Page Up) command. This isvery handy when you want to translatetwo source sentences as one target

sentence or when TRADOS does notrecognize a period at the end of a sen-tence as a segmentation mark (becausethere is a number or unknown abbrevi-ation right before the period). In suchsituations, the segment that is openedfor translation consists of two sen-tences. You can expand or shrink asegment by repeatedly selecting therespective command as many times asyou need to, and the segmentexpands/shrinks each time to thenext/previous segmentation mark.

In Love With the New TRADOSMultiTerm

If you got frustrated withMultiTerm Version 5 and have nottried it since, you should take a look atthe newer version of iX or 7. The newversions are much easier to use, andthey have made it very simple to addnew terms to a MultiTerm glossarydirectly from Word or TagEditorwhile translating. I use this often toadd long product names to a glos-sary—not so much to build a long-term glossary, but to make it easy tocopy these long names into my trans-lation with a simple “Alt+Arrow”keyboard command without having toretype the entire name.

Another handy idea is to maintain atermbase of common long words andexpressions, such as “turn right,” “coun-terclockwise,” and “immediately.” Thiscould be used for the same purpose asMS Word’s AutoCorrect function (seethe article by Clifford Landers in theJune 2006 issue of The ATA Chronicle,pages 26-28), but the nice thing aboutthis is that you do not have to memorizeanything, because MultiTerm remem-bers the terms for you.

Converting your old MultiTerm 5,Word, or Excel glossaries toMultiTerm 7 termbases is still a muchmore complicated process than itshould be. However, it is not dif- ➡

Figure 2. Selecting project and filter settings for atranslation project before starting a translation. Selectedfields and their values are listed in the “Current settings”box.


ficult if you know what to do. In caseyou do not, you can download myQuick Guide to MultiTerm iX/7 fromwww.finntranslations.com/tradoshelp.html. It includes simple, straight-forward instructions on how to con-vert and exchange glossaries and howto add terms directly from Word andTagEditor to a MultiTerm termbase(i.e., the basics you need to know inorder to use MultiTerm efficientlywith TRADOS).

SubstitutionsIn addition to the default substitu-

tion of numbers, dates, and measure-ments, TRADOS also offers twonon-default substitution possibilities(see Figure 4).

Acronyms Substitution: The “Acro-nyms Substitution” setting (File >Setup… > Substitutions > Acronyms)becomes handy when you areworking, for example, on software orhardware interface instructions wheresoftware commands or names of but-tons and switches are written all incapital letters and do not need to betranslated. When “Acronyms” isselected on the “Substitutions” tab,TRADOS treats all words written incaps as placeables, and substitutesthem automatically. For example,“Select SEGMENTATION RULESTab” would be a 100% match after“Select ACCESS RIGHTS Tab” hasbeen translated.

Variables Substitution: The othersubstitution option that everyoneshould be using is called “Variables”(File > Setup… > Substitutions >Variables). With this you can definewords that do not need to be translated,such as proper names, by adding themto the “Variable List” box. All of thesewords will then be treated as place-ables during translation, and it is very

Figure 3. Segmentation Rules tab open for adjusting segmentation rules. Here, a comma has been added as a “stopcharacter” in the “Marks” category to allow translation from comma to comma.


Figure 4. Substitutions tab open for selecting automatic substitutions. Variable names in the “Variable list” will betreated as placeables during translation.




easy to add them to the text using the“Add Placeable” commands withoutever again having to retype them.

Not Fuzzy EnoughYou just translated a similar sen-

tence or phrase, but cannot figure outwhy it will not come up as a fuzzymatch. Try lowering your “MinimumMatch Value %” setting from thedefault 70% down to 50% or 40%.These lower level fuzzy matches ofteninclude useful phrases and terms. Youwill also get some useless matches, butthey do not cost you anything—youcan just ignore them. To change thesetting, select “Options > TranslationMemory Options > General.”

Using Reference TM forConcordance Searches

As I mentioned earlier, TRADOSallows you to have only one TM openat a time for translation, but you canuse a second memory for concordancesearches. Reference (read-only)memory for concordance is a handyway to utilize a second TM. You canset TRADOS so that every concor-dance search is also done in the refer-ence translation memory in addition tothe active translation memory. If youselect “Search First in ReferenceTranslation Memory,” the search isdone first in the reference TM and thenin the active TM (if there were no hitsin the reference TM). Otherwise, thesearch is done first in the active TM. Toselect the reference TM and change thesetting, select “Options > TranslationMemory Options > Concordance.”

Microsoft GlossariesIf you have been reading Jost

Zetzsche’s excellent Tool Kit newsletter(www.internationalwriters.com), youknow about the new arrangement withMicrosoft glossaries. The application-specific “glossaries,” which were more

like translation memories, have beenreplaced by a real glossary of 9,000English terms, plus their translations inup to 45 different languages. The pre-vious glossary files are now availableonly to paying subscribers, so do notlose your old files if you still havethem.

I have often been asked how to useMicrosoft glossaries with TRADOS andMultiTerm. Here are some suggestions.

First, convert those old glossaryfiles into a TRADOS TM. One simpleway to do it is through a free utilitycalled MSGloss2TWB (www.global-ready. com). Depending on the lan-guage, this memory can havehundreds of thousands of TUs and canbe very useful if you translate any soft-

ware-related texts. This is one of the fewmemories I keep separate from my mainTM, and I use it often as a referencememory for concordance searches.

Unfortunately, the MSGloss2TWBprogram does not have an option toinclude attribute fields in the resultingmemory, but you can add them quiteeasily in TRADOS using the “Findand Replace” function.

1. First, you need to set up the fieldsand their values if they do notalready exist in the memory, asdescribed earlier (see Figure 1).

2. Go to “File > Maintenance…>Find and Replace > More,” selectthe attribute field(s) and the ➡

Useful Links

www.apsic.comApSIC Xbench: for searching and viewing bilingual files, such as MS glossaries.ApSIC Comparator: for comparing edited and unedited files, such as TagEditor files.

www.finntranslations.com/tradoshelp.htmlQuick Guide to MultiTerm iX/7, along with other workshop and presentation handouts.

www.globalready.comMSGloss2TWB for converting MS glossaries to TRADOS TM.

www.internationalwriters.comJost Zetzsche’s Tool Kit newsletter.

www.microsoft.com/globaldev/tools/MILSGlossary.mspxThe new Microsoft terminology translations link.

http://sdl.custhelp.com The SDL/TRADOS support page.



value(s) you want to add from thepicklist(s).

3. Select “Merge” and click “OK.”

4. Select “Begin Search” to view thefirst few TUs that will be changed.

5. Select “Change” and “AllTranslation Units.”

If you want to test your find andreplace operation, select the“Translation Units in Current List”option instead. That way the changewill only be applied to those TUs inthe list, and you can verify the effectof the operation before applying it tothe entire memory. You can also dothis with a set of proper find andreplace commands in a good texteditor (such as UltraEdit), or even inWord (Notepad is too slow for this)before importing the .txt files to a TM.

Another handy and equally cheapway to utilize these files is anotherfree utility called ApSIC Xbench

(www.apsic.com). It does not allowyou to import the files into aTRADOS TM, but it provides a pow-erful and convenient way to searchand view bilingual information, suchas the old Microsoft glossary files,outside of TRADOS.

The old Microsoft glossary filesare not suitable for use in MultiTermbecause they mostly consist of wholesentences or paragraphs of text.MultiTerm needs a proper glossary ofterms or short phrases. Years ago, theMicrosoft glossary package used toinclude a “Glossary of StandardTerms” of over a thousand generalterms, which made it a very usefulsource for a MultiTerm glossary.However, this file has not been a partof the package for many years. Maybeyou still have it from some old down-load. The file name is xx_stand.csv(“xx” indicates a language code).Anyhow, the new free 9,000-termglossary file also suits this purposevery well, but you might first want todelete all the languages you do not

need to make the conversion faster.The conversion is done withMultiTerm Convert and by selectingExcel as the source file type. You canfind detailed instructions in my QuickGuide to MultiTerm iX/7.

In and Out of the WoodsI hope these few suggestions and

examples have given you the courageand knowledge to venture deeper intothe TRADOS environment and startpicking even juicier and biggerberries. If, in spite of it all, you end upgetting lost, remember that everyTRADOS dialog box has a “Help”button that opens online help for thatparticular dialog box. Just stay calmand read the instructions. The bears?Yes, they are out there, but do notworry. Just play dead, or are you sup-posed to run away, or is it act big andaggressive…? Hmm, I need to checkon that.


Plan now to exhibit at ATA’s 48th Annual Conference in San Francisco,California, October 31-November 3, 2007. Exhibiting at ATA’s Annual

Conference offers the best opportunity to market your products and servicesface-to-face to more than 1,500 translators and interpreters in one

location. Translators and interpreters are consumers of computer hardware and software, technical publications and reference

books, office products, and much more.

For additional information, please contact Matt Hicks,McNeill Group Inc.; [email protected];

(215) 321-9662, ext. 19; Fax: (215) 321-9636.

Attention Exhibitors ATA’s 48th Annual Conference • San Francisco, California • October 31-November 3, 2007


Documents

CHRONICLE - Finnish Translation Services and Trados Training