13.07.2015 Views

A comparative study of terminology management tools in ... - Transsoft

A comparative study of terminology management tools in ... - Transsoft

A comparative study of terminology management tools in ... - Transsoft

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

CHAPTER IV....................................................................................................................................721. INTRODUCTION......................................................................................................................722. THE RESULT SHEET..............................................................................................................723. COMMENTS..............................................................................................................................974. CONCLUSIONS.........................................................................................................................98CONCLUSION................................................................................................................................100BIBLIOGRAPHY:...........................................................................................................................101APPENDIX I....................................................................................................................................105APPENDIX II...................................................................................................................................109APPENDIX III.................................................................................................................................1113


ACKNOWLEDGEMENTSI would like to express my gratitude to Marc<strong>in</strong> Feder Ph.D. from Adam MickiewiczUniversity for arous<strong>in</strong>g my <strong>in</strong>terest <strong>in</strong> <strong>term<strong>in</strong>ology</strong> and computer-assisted translation, as wellas for provid<strong>in</strong>g his illum<strong>in</strong>at<strong>in</strong>g comments and <strong>in</strong>valuable advice. I would also like to thankPeter Sandr<strong>in</strong>i PhD from the University <strong>of</strong> Innsbruck for provid<strong>in</strong>g me with materials whichgreatly contributed to the development <strong>of</strong> my knowledge <strong>of</strong> <strong>term<strong>in</strong>ology</strong> <strong>management</strong> andwhich po<strong>in</strong>ted me <strong>in</strong> the right direction <strong>in</strong> my research. Last but not least, I would like tothank my supervisor for his encouragement and immense patience <strong>in</strong> support<strong>in</strong>g me at allstages <strong>of</strong> my work.4


INTRODUCTIONIn the global village we live <strong>in</strong>, the need to communicate seamlessly and effectivelyis a very significant one. The translation market <strong>in</strong> Poland is nowadays becom<strong>in</strong>g highlycompetitive, demand<strong>in</strong>g ever higher standards <strong>of</strong> performance and productivity both fromexperienced and novice translators. The requirements Polish translators and <strong>in</strong>terpreters willhave to meet are bound to soar even higher now that Poland has become a member <strong>of</strong> theEuropean Union.The Polish translation market can be characterized as highly fragmented, with veryfew medium-size companies. Many translators run s<strong>in</strong>gle-person companies seated <strong>in</strong> theirhomes, sometimes not even own<strong>in</strong>g legal s<strong>of</strong>tware. The fact that the operat<strong>in</strong>g costs <strong>of</strong> thesetranslation agencies are low, frequently paired with excellent customer service and highoutput quality can be viewed as an advantage. However, there are a number <strong>of</strong> seriouslimitations which act to the detriment <strong>of</strong> translation capacity and quality <strong>of</strong> the Polishtranslation agencies when it comes to large translation projects or technically demand<strong>in</strong>gassignments (Argos 2002). Good <strong>management</strong> <strong>of</strong> translation projects, which <strong>in</strong>volves,among others, efficient <strong>term<strong>in</strong>ology</strong> <strong>management</strong> as well as the ability to use and benefitfrom the state-<strong>of</strong>-the-art language technology has become a necessity for those who wish torema<strong>in</strong> on the market.Therefore, there is a great need for comprehensive writ<strong>in</strong>g on the <strong>tools</strong> that mighthelp translators meet the ever-<strong>in</strong>creas<strong>in</strong>g expectations <strong>of</strong> their clients. This need <strong>in</strong>cludes notonly a comprehensive presentation <strong>of</strong> the <strong>tools</strong>, their functionalities and advertis<strong>in</strong>g thedifferent applications available, but also detailed and objective guidel<strong>in</strong>es on how to evaluatesuch <strong>tools</strong>. There is a great abundance <strong>of</strong> sources present<strong>in</strong>g CAT <strong>tools</strong>. However, onlytest<strong>in</strong>g the <strong>tools</strong> aga<strong>in</strong>st objective and comprehensive criteria can give a real picture <strong>of</strong> the<strong>tools</strong>’ applicability for a given user or a particular work<strong>in</strong>g environment. There is also animmense need to promote standards and new developments <strong>in</strong> the areas <strong>of</strong> languageeng<strong>in</strong>eer<strong>in</strong>g and general computer technologies, <strong>in</strong> order to ensure more compatibility andexchangeability <strong>of</strong> <strong>term<strong>in</strong>ology</strong> resources and translation memories among translators,technical writers, etc. (POINTER 1996) 1 .Bear<strong>in</strong>g <strong>in</strong> m<strong>in</strong>d the current situation on the Polish and global translation market, theauthor decided to devote this thesis to present<strong>in</strong>g a comprehensive <strong>study</strong> <strong>of</strong> <strong>term<strong>in</strong>ology</strong>1 POINTER (Proposals for an Operational Infrastructure for Term<strong>in</strong>ology <strong>in</strong> Europe) - a project carried out by<strong>term<strong>in</strong>ology</strong> specialists <strong>in</strong> the years 1995 - 19965


<strong>management</strong> <strong>tools</strong>, which may address the above-mentioned needs, although <strong>in</strong> the limitedway. The <strong>study</strong> will <strong>in</strong>volve theoretical <strong>in</strong>troduction <strong>in</strong>to <strong>term<strong>in</strong>ology</strong> <strong>management</strong> as ascientific discipl<strong>in</strong>e <strong>in</strong>clud<strong>in</strong>g a brief historical outl<strong>in</strong>e, followed by a suggestedmethodology <strong>of</strong> evaluation and an exemplary evaluation procedure compar<strong>in</strong>g three<strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong>. The author would like to emphasize that all the <strong>tools</strong>selected for presentation <strong>in</strong> the thesis were available for test<strong>in</strong>g for free, and no s<strong>of</strong>twareprovider sponsored this project. The reason<strong>in</strong>g beh<strong>in</strong>d the selection <strong>of</strong> particular <strong>tools</strong> isgiven <strong>in</strong> the fourth chapter <strong>of</strong> the thesis.In order to demonstrate the features <strong>of</strong> the <strong>tools</strong>, term<strong>in</strong>ological databases werecreated <strong>in</strong> the programs selected. As a corpus for <strong>term<strong>in</strong>ology</strong> extraction for the termbases,the author used a number <strong>of</strong> sources (see Appendix I). The selected programs were <strong>in</strong>stalledon two computers, both hav<strong>in</strong>g W<strong>in</strong>dows XP for the operat<strong>in</strong>g system. One had 256 MBRAM, the other 128 MB RAM. In both cases Office 2003 was used.6


CHAPTER ITERMINOLOGY – BASIC CONCEPTS1. INTRODUCTIONThis thesis conta<strong>in</strong>s four chapters, each devoted to a different aspect <strong>of</strong> the <strong>study</strong> <strong>of</strong><strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong>. This chapter acts as an <strong>in</strong>troduction to <strong>term<strong>in</strong>ology</strong> and<strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>in</strong> general. First the readers will be presented with the def<strong>in</strong>itions<strong>of</strong> basic concepts <strong>of</strong> this discipl<strong>in</strong>e. The historical background <strong>of</strong> <strong>term<strong>in</strong>ology</strong> <strong>management</strong>will <strong>in</strong>troduce the most important developments <strong>in</strong> this branch <strong>of</strong> knowledge. The f<strong>in</strong>al part<strong>of</strong> this chapter will be devoted to illustrat<strong>in</strong>g the significance <strong>of</strong> <strong>term<strong>in</strong>ology</strong> <strong>management</strong>and <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> <strong>in</strong> translation.2. DEFINITIONSThe notions central to <strong>term<strong>in</strong>ology</strong> <strong>management</strong>, and thus to this thesis, are term andconcept. As def<strong>in</strong>ed by Trippel “term is the language sign for a concept. This language signdoes not necessarily have to be a s<strong>in</strong>gle word, but it can also be a set <strong>of</strong> words - a fixedphrase - used only to denote a specific concept. Terms are not language <strong>in</strong>dependent whileconcepts are.”(Trippel 1999).The term <strong>term<strong>in</strong>ology</strong> has two possible <strong>in</strong>terpretations. The first one says that it is aspecialist vocabulary, used <strong>in</strong> a particular subject field, also referred to as technical jargon.The other reference <strong>of</strong> this term is the theory or science deal<strong>in</strong>g with the relations betweenterms and concepts (Trippel 1999). Another def<strong>in</strong>ition states that it is ‘a structured set <strong>of</strong>concepts and their designations (graphical symbols, terms, phraseological units, etc.) <strong>in</strong> aspecific subject field.’ (POINTER 1996). On the whole, it is an <strong>in</strong>terdiscipl<strong>in</strong>ary branchwhich <strong>in</strong>volves both theoretical and practical aspects <strong>of</strong> creation, <strong>in</strong>troduction, <strong>in</strong>terpretation,usage, validation, evaluation, correction and classification <strong>of</strong> terms. There are a number <strong>of</strong>applications <strong>of</strong> <strong>term<strong>in</strong>ology</strong>, among which the most significant are: standardization, researchand development, market<strong>in</strong>g communications, consumer <strong>in</strong>formation, language eng<strong>in</strong>eer<strong>in</strong>gapplications, knowledge eng<strong>in</strong>eer<strong>in</strong>g, computer-aided language learn<strong>in</strong>g (CALL), distancelearn<strong>in</strong>g, computer-aided <strong>in</strong>struction (CAI), technical writ<strong>in</strong>g, corporate <strong>in</strong>formation7


systems, <strong>in</strong>formation retrieval, term databanks (TDB), computer-aided translation (CAT),mach<strong>in</strong>e translation (MT), human translation, and nomenclature (POINTER 1996).It is <strong>in</strong>structive to draw a dist<strong>in</strong>ction between the seem<strong>in</strong>gly similar discipl<strong>in</strong>es <strong>of</strong><strong>term<strong>in</strong>ology</strong> and lexicology, as well as term<strong>in</strong>ography and lexicography. While themethodology <strong>of</strong> the discipl<strong>in</strong>es <strong>in</strong> question may be <strong>in</strong> some cases similar, their focus isdifferent. Lexicology is a l<strong>in</strong>guistic specialty deal<strong>in</strong>g with general language vocabulary,while <strong>term<strong>in</strong>ology</strong> deals exclusively with special language lexis (POINTER 1996).Similarly, the general language dictionaries, compiled as a result <strong>of</strong> lexicographical work,conta<strong>in</strong> some specialist terms as part <strong>of</strong> the general vocabulary, however usually embedded<strong>in</strong> the general language entries. Term<strong>in</strong>ography <strong>in</strong> turn, deals with compil<strong>in</strong>g speciallanguage vocabulary collections solely (POINTER 1996). Another difference is manifested<strong>in</strong> the direction <strong>of</strong> work. Term<strong>in</strong>ology collection, usually restricted to a specialist doma<strong>in</strong>,beg<strong>in</strong>s with concepts, not terms themselves and proceeds with the mapp<strong>in</strong>g <strong>of</strong> the doma<strong>in</strong>with the concept delimitations i.e. terms, whereas lexicography work starts with vocabularycollection. However, there are l<strong>in</strong>guists who claim that the dist<strong>in</strong>ction between the twodiscipl<strong>in</strong>es may soon be no longer valid due to the imm<strong>in</strong>ent convergence <strong>of</strong> theirmethodologies (Campenhoudt 2001).In this thesis the author will focus on the practical aspects <strong>of</strong> <strong>term<strong>in</strong>ology</strong>, and itsapplication <strong>in</strong> mach<strong>in</strong>e-assisted human translation (MAHT), therefore only selected issuesconnected with <strong>term<strong>in</strong>ology</strong> <strong>management</strong> will be discussed.Term<strong>in</strong>ology <strong>management</strong> <strong>in</strong>volves a number <strong>of</strong> activities, rang<strong>in</strong>g from <strong>term<strong>in</strong>ology</strong>collection or extraction, to <strong>term<strong>in</strong>ology</strong> creation and validation, to classification, storage,retrieval and exchange. For the purposes <strong>of</strong> this thesis, we will focus only on the follow<strong>in</strong>gaspects <strong>of</strong> <strong>term<strong>in</strong>ology</strong> <strong>management</strong>: <strong>term<strong>in</strong>ology</strong> extraction, organization, storage, retrievaland exchange. Some aspects <strong>of</strong> validation will also be mentioned.This thesis is devoted to discuss<strong>in</strong>g <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> which are <strong>of</strong>tenreferred to as <strong>term<strong>in</strong>ology</strong> <strong>management</strong> systems (TMS). They are s<strong>of</strong>tware systems whichhelp to create and store term<strong>in</strong>ological data <strong>in</strong> the form which allows for a controlled use <strong>of</strong>the data. Term<strong>in</strong>ology <strong>management</strong> systems have nowadays become <strong>in</strong>dispensable <strong>tools</strong> fortranslation agencies and translation project managers. Thus, at least rudimentary knowledge<strong>of</strong> these systems is required <strong>of</strong> translators who seek employment with such agencies.Another central notion is that <strong>of</strong> term<strong>in</strong>ological database, or termbase:8


‘Termbase: Short form <strong>of</strong> Term<strong>in</strong>ology database. A termbase is thecollection <strong>of</strong> <strong>in</strong>formation on a term or concept <strong>in</strong> a structured,electronically readable way comb<strong>in</strong>ed with a <strong>term<strong>in</strong>ology</strong> <strong>management</strong>system. It is mostly used synonymously with termbank, though someterm<strong>in</strong>ologists dist<strong>in</strong>guish them. If they are dist<strong>in</strong>guished, <strong>term<strong>in</strong>ology</strong>databases do not <strong>in</strong>clude the organizational environment but termbanksdo.’ (Trippel 1999) (cf. Gal<strong>in</strong>ski 1998).In this thesis the terms termbank and termbase will be used <strong>in</strong>terchangeably.Term<strong>in</strong>ology <strong>management</strong> <strong>tools</strong> are part <strong>of</strong> a larger group <strong>of</strong> s<strong>of</strong>tware <strong>tools</strong> referredto as computer-assisted translation (CAT) <strong>tools</strong>. CAT is def<strong>in</strong>ed as ‘direct translation byhumans with the help <strong>of</strong> a computer <strong>in</strong>terface which makes translational expertise accessiblethrough “translation-<strong>in</strong>telligent” s<strong>of</strong>tware’. (Neubert 1991:56). In other words, CATapplications are a group <strong>of</strong> s<strong>of</strong>tware <strong>tools</strong> assist<strong>in</strong>g translators, where the human knowledgeand l<strong>in</strong>guistic competence are the key factors, and it is the human translator who plays thedom<strong>in</strong>ant role and makes the f<strong>in</strong>al decisions concern<strong>in</strong>g <strong>term<strong>in</strong>ology</strong> and phraseologychoices.Modern CAT <strong>tools</strong>, referred to as workbenches, consist <strong>of</strong> a number <strong>of</strong> modules orcomponents, <strong>term<strong>in</strong>ology</strong> <strong>management</strong> systems be<strong>in</strong>g part <strong>of</strong> them. The module which isconsidered the central one though, is the translation memory module.‘There are different TM programs currently available on the market,but they share similar features, albeit with some differences <strong>in</strong> speed and data<strong>management</strong>. Normally, the core <strong>of</strong> TM is the memory, a complex databasewhere source text sentences are aligned side by side with the correspond<strong>in</strong>gtarget text sentences. The ways <strong>in</strong> which the memory can be accessed andmanaged vary from one TM program to the other, but the philosophy beh<strong>in</strong>dthe tool is basically the same: reus<strong>in</strong>g previous work.’(Rico Pérez 2001).In a nutshell, TM <strong>tools</strong> play the role <strong>of</strong> a perfect memory that can be accessedanytime dur<strong>in</strong>g the translation process. It is a memory that never fails to retrieve therequested <strong>in</strong>formation and prevents the translator from struggl<strong>in</strong>g with the same translationproblem twice. The fact that translation memory stores aligned sentence pairs <strong>in</strong> sourcelanguage (SL) and target language (TL) makes the tool extremely useful <strong>in</strong> translat<strong>in</strong>g9


epetitive texts e.g. technical manuals. When a new document is be<strong>in</strong>g translated <strong>in</strong> aworkbench environment, the program automatically searches the translation memory foridentical or similar segments, and whenever a match is returned (exact or fuzzy) it will bedisplayed <strong>in</strong> a special pane or grid or directly <strong>in</strong> the space where the target segment shouldbe entered.There are, however, technical texts that are very dense <strong>in</strong> terms <strong>of</strong> specialist<strong>term<strong>in</strong>ology</strong>, but do not conta<strong>in</strong> as much as two identical sentences. In this case the<strong>term<strong>in</strong>ology</strong> <strong>management</strong> component <strong>of</strong> a workbench comes <strong>in</strong> as the right solution (Benis1998). Thanks to <strong>term<strong>in</strong>ology</strong> <strong>management</strong> modules even <strong>in</strong> the case <strong>of</strong> non-repetitive textswe can still benefit from the workbench packages <strong>in</strong> terms <strong>of</strong> speed and quality <strong>of</strong>translation, even though translation memory is not applicable.Other components which are normally part <strong>of</strong> workbench applications are alignment<strong>tools</strong> (applications used for build<strong>in</strong>g translation memories from the correspond<strong>in</strong>g SL andTL documents), analysis modules perform<strong>in</strong>g word frequency and repeatability calculations,sometimes also database and project ma<strong>in</strong>tenance modules.At this stage it is necessary to draw the dist<strong>in</strong>ction between two terms which arefrequently confused, i.e. computer-assisted translation (CAT), also referred to as mach<strong>in</strong>eassistedhuman translation (MAHT), and mach<strong>in</strong>e translation (MT). While it clearlytranspires from the very term that MAHT is the type <strong>of</strong> translation where the humantranslator plays the crucial role (cf. Feder 2001:51, Neubert 1991:57) it should be noted that‘MT aims at assembl<strong>in</strong>g all the <strong>in</strong>formation necessary for translation <strong>in</strong> one program so thata text can be translated without human <strong>in</strong>tervention’ (Craciunescu et al. 2004). Thedifference between MAHT and MT applications is also <strong>in</strong> the output quality. In the case <strong>of</strong>MAHT <strong>tools</strong>, the translations are usually <strong>of</strong> publish<strong>in</strong>g quality. The up-to-date MT systems,on the other hand, deliver translations <strong>of</strong> unacceptable quality or requir<strong>in</strong>g much postedit<strong>in</strong>g.However, the advent <strong>of</strong> new MT systems apply<strong>in</strong>g neural networks and artificial<strong>in</strong>telligence technology is only a matter <strong>of</strong> time and we may expect the quality <strong>of</strong> theiroutput to improve (Champollion: 2001).F<strong>in</strong>ally, we should bear <strong>in</strong> m<strong>in</strong>d that computer-assisted translation is a complexprocess consist<strong>in</strong>g <strong>of</strong> a several stages. Manag<strong>in</strong>g large translation projects <strong>in</strong>volves a number<strong>of</strong> phases and tasks which can be broken down <strong>in</strong>to translation and non-translation tasks orpre-translation, translation and post-translation tasks 2 . In such classifications, only the actual2 http://www.ad-ex.net/process.pdf10


uild<strong>in</strong>g <strong>of</strong> the TL equivalent <strong>of</strong> a SL text is regarded as a translation task, while all therema<strong>in</strong><strong>in</strong>g tasks, i.e. <strong>term<strong>in</strong>ology</strong> <strong>management</strong>, desktop publish<strong>in</strong>g (format conversions), textextraction 3 , pro<strong>of</strong>read<strong>in</strong>g and customer’s review, are considered to be non-translation tasks.3. HISTORICAL BACKGROUNDThe first efforts <strong>in</strong> <strong>term<strong>in</strong>ology</strong> began <strong>in</strong> 1960’s, probably as a result <strong>of</strong> thepublication <strong>of</strong> the (<strong>in</strong>)famous ALPAC report which advocated develop<strong>in</strong>g s<strong>of</strong>tware <strong>tools</strong> toaid translators <strong>in</strong>stead <strong>of</strong> carry<strong>in</strong>g out mach<strong>in</strong>e translation research (Feder 2001:15, Hutch<strong>in</strong>s1996, Palacz 2003:8). Term<strong>in</strong>ology, however, had not been perceived as a discipl<strong>in</strong>e dist<strong>in</strong>ctfrom lexicology and other l<strong>in</strong>guistic discipl<strong>in</strong>es until the publication <strong>of</strong> the E<strong>in</strong>führung <strong>in</strong> dieallgeme<strong>in</strong>e Term<strong>in</strong>ologielehre und term<strong>in</strong>ologische Lexikographie 4 by Eugen Wüster <strong>in</strong>1979 (Wüster 1979) (cf. Campenhoudt 2001).The first <strong>term<strong>in</strong>ology</strong> projects were only available for large organizations because the<strong>term<strong>in</strong>ology</strong> <strong>management</strong> s<strong>of</strong>tware required ma<strong>in</strong>frame computers (Rico Pérez 2001). Thissituation led to the development <strong>of</strong> large-scale termbanks, e.g. Termium, Eurodicautom,Banque de term<strong>in</strong>ologie du Quebec (now Le Grande dictionnaire term<strong>in</strong>ologique). Thetermbanks developed at this stage are still <strong>in</strong> use nowadays, although the systems underwentgeneral overhauls, e.g. Eurodicautom now runs on entirely new platforms (Oracle andFulcrum) 5 .The 1980’s saw the first electronic dictionaries and <strong>term<strong>in</strong>ology</strong> <strong>management</strong>s<strong>of</strong>tware developed for personal computers and available for <strong>in</strong>dividual translators,follow<strong>in</strong>g the development <strong>of</strong> translation memory s<strong>of</strong>tware. However, these <strong>tools</strong> had manylimitations. First <strong>of</strong> all, they were not networkable i.e. it was impossible to share <strong>term<strong>in</strong>ology</strong>collections over local area networks. The first generation <strong>of</strong> <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong><strong>of</strong>ten <strong>of</strong>fered only unidirectional search<strong>in</strong>g e.g. EN-GR but not GR-EN. There were alsorestrictions on the number and type <strong>of</strong> data fields, as well as <strong>of</strong> storage capacity.The new generation <strong>of</strong> <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> followed the publication <strong>of</strong> theconcepts <strong>of</strong> a three-level <strong>in</strong>tegrated translator’s workstation (Melby 1992) (cf. Feder 2001).3 If a project manager does not receive a source file <strong>in</strong> a text format, e.g. the source file is <strong>in</strong> a protected PDFformat, the text that is to be translated will <strong>of</strong>ten be extracted from the file and delivered to the translator <strong>in</strong> atext format, <strong>in</strong> order to enable the translator to use CAT s<strong>of</strong>tware. If the source text is delivered <strong>in</strong> hard copy,text extraction will refer to us<strong>in</strong>g optical character recognition (OCR) <strong>in</strong> order to receive the source text <strong>in</strong> theelectronic form.4 General Theory <strong>of</strong> Term<strong>in</strong>ology and Term<strong>in</strong>ological Lexicography – An Introduction5 http://europa.eu.<strong>in</strong>t/eurodicautom/Controller?ACTION=about11


The first release <strong>of</strong> MultiTerm for W<strong>in</strong>dows and DOS Translator’s Workbench package was<strong>in</strong> 1992. Another CAT tool which is now one <strong>of</strong> the market leaders – Déjà Vu, was firstreleased <strong>in</strong> 1993. These <strong>tools</strong> <strong>of</strong>fered more possibilities than the earlier generation andga<strong>in</strong>ed significance especially due to the fact that the developers e.g. Trados, recognized theopportunities ly<strong>in</strong>g <strong>in</strong> the grow<strong>in</strong>g popularity <strong>of</strong> local area networks (Brace&Joscelyne1994).S<strong>in</strong>ce that time, many new <strong>tools</strong> and new versions <strong>of</strong> the first CAT <strong>tools</strong> have beenreleased, catch<strong>in</strong>g up with the developments <strong>in</strong> language eng<strong>in</strong>eer<strong>in</strong>g and computertechnology. Currently, there are two ma<strong>in</strong> tendencies <strong>in</strong> CAT development. On the one hand,s<strong>of</strong>tware developers tend to isolate the functionalities which used to be part <strong>of</strong> <strong>term<strong>in</strong>ology</strong>modules <strong>in</strong>to separate <strong>tools</strong>, e.g. term extraction module was part <strong>of</strong> MultiTerm 5.0 but is nolonger <strong>in</strong> a package with Multiterm iX. On the other hand, there is the tendency for<strong>in</strong>tegration <strong>of</strong> the typical MAHT <strong>tools</strong>, <strong>in</strong>clud<strong>in</strong>g <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong>, withmach<strong>in</strong>e translation systems and localization <strong>tools</strong> (Melby&Wright 1999), result<strong>in</strong>g <strong>in</strong> theso-called hybrid systems (Feder 2001:32). The application <strong>of</strong> hybrid systems and highly<strong>in</strong>tegrated translation environment is usually most advanced <strong>in</strong> large <strong>in</strong>stitutions, e.g.European Commission (Blatt 1998, Hutch<strong>in</strong>s 1989).4. SIGNIFICANCE OF TERMINOLOGY MANAGEMENT FORTRANSLATION:After the <strong>in</strong>troduction <strong>of</strong> the basic notions related to <strong>term<strong>in</strong>ology</strong> we should nowfocus on how <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> can assist translators <strong>in</strong> their work. The bestanswer to this question is provided by translators themselves:‘Why do translators need to consult dictionaries, databases and/or expertswhen they work? The answer is so obvious that we tend to forget howimportant it is: Translators are not experts. This fact colors our wholeapproach to our work, particularly <strong>in</strong> areas where we are less thanconfident <strong>of</strong> our mastery <strong>of</strong> the subject matter.’ (Titchen&Fraser 1996)As we can see, applications designed to create and facilitate the use <strong>of</strong> specialist referencesources, tailored to the needs <strong>of</strong> translators cannot be replaced by any other <strong>tools</strong>. For atranslator who has no knowledge <strong>of</strong> a particular subject area and needs a number <strong>of</strong> technical12


terms which cannot be found <strong>in</strong> general-language dictionaries there is no alternative, but tocreate reliable <strong>term<strong>in</strong>ology</strong> collections themselves. Another l<strong>in</strong>guist mentions the follow<strong>in</strong>gargument <strong>in</strong> support <strong>of</strong> termbases <strong>in</strong> translation work:‘(…) <strong>in</strong> his daily work rout<strong>in</strong>e even the experienced translatorencounters countless “new” problems hav<strong>in</strong>g to do with the almost unlimited<strong>in</strong>flux <strong>of</strong> words, terms and phrases that are not part <strong>of</strong> his average or evenspecialist lexical knowledge. The way he successfully copes with these gapscan most efficiently be modeled by term banks on the computer.’ (Neubert1991:58)One more reason for the emphasis placed nowadays on efficient <strong>term<strong>in</strong>ology</strong><strong>management</strong> employ<strong>in</strong>g state-<strong>of</strong>-the-art technologies is the impact <strong>of</strong> <strong>term<strong>in</strong>ology</strong> used, onthe localization market, i.e. one <strong>of</strong> the most important areas <strong>of</strong> technological development.‘Efficient <strong>term<strong>in</strong>ology</strong> <strong>management</strong> is crucial for publishers andmanufacturers when translat<strong>in</strong>g and localiz<strong>in</strong>g their products. Translationvendors and translators may change over time – but the quality <strong>of</strong> thelocalized product should always adhere to the highest possible standard.Consistent <strong>term<strong>in</strong>ology</strong> is necessary for ensur<strong>in</strong>g cont<strong>in</strong>ued familiarity witha product, and it is essential for functional compatibility between differentversions <strong>of</strong> a product on one or multiple platforms.’ (Project Review 2000)Another advantage <strong>of</strong> us<strong>in</strong>g <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> <strong>in</strong> translation is that asearch for a given term is more time-efficient when compared to search<strong>in</strong>g <strong>in</strong> pr<strong>in</strong>teddictionaries and other sources. It is not only faster however, but also more reliable as it waspo<strong>in</strong>ted out by the specialists work<strong>in</strong>g on the POINTER project <strong>in</strong> 1995 and 1996:‘Analysis <strong>of</strong> various dictionary entries demonstrates that the extraction <strong>of</strong>term<strong>in</strong>ological data from currently-available LGP 6 dictionaries (bothmonol<strong>in</strong>gual and bil<strong>in</strong>gual) is problematic from a number <strong>of</strong> different po<strong>in</strong>ts<strong>of</strong> view, <strong>in</strong>clud<strong>in</strong>g the <strong>in</strong>consistent and imprecise use <strong>of</strong> subject-field labels,6 Language for General Purposes13


the absence <strong>of</strong> adequate pragmatic <strong>in</strong>formation, and vary<strong>in</strong>g def<strong>in</strong>itionalpractices. Terms are also <strong>of</strong>ten deeply nested <strong>in</strong> entries, even as sub-senses <strong>of</strong>polysemous headwords. The unsatisfactory use <strong>of</strong> subject-field labels is <strong>of</strong>particular importance for the automatic extraction <strong>of</strong> data.’ (POINTER 1996)Term<strong>in</strong>ological databases are designed to avoid problems <strong>of</strong> <strong>in</strong>consistency and imprecision<strong>of</strong> LGP dictionaries. Term<strong>in</strong>ological records provided <strong>in</strong> termbases are prepared bytranslators on the basis <strong>of</strong> sources they trust, with usage contexts <strong>of</strong> native orig<strong>in</strong> exclusively(Göpferich 1995:23) frequently validated follow<strong>in</strong>g consultations with experts <strong>in</strong> givensubject areas. Thus, the reliability <strong>of</strong> l<strong>in</strong>guistic data <strong>in</strong>cluded is much higher.One more argument <strong>in</strong> favor <strong>of</strong> us<strong>in</strong>g <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> <strong>in</strong> translation isthat usually there are no specialist dictionaries <strong>in</strong> the new and quickly develop<strong>in</strong>g fields <strong>of</strong>knowledge or ‘the production <strong>of</strong> up-to-date reference works is lagg<strong>in</strong>g beh<strong>in</strong>d (Špela: 2001).The reason for this situation is that the process <strong>of</strong> compilation and publish<strong>in</strong>g <strong>of</strong> pr<strong>in</strong>teddictionaries takes much longer and is more costly than <strong>in</strong> the case <strong>of</strong> electronic <strong>term<strong>in</strong>ology</strong>collections. Therefore, mach<strong>in</strong>e-readable sources can reach the users much faster. Moreover,it is much easier to update and modify an electronic termbase than a pr<strong>in</strong>ted dictionary.Another obvious advantage, perhaps the most significant one, is that the results <strong>of</strong><strong>term<strong>in</strong>ology</strong> research once carried out are saved and kept for reuse <strong>in</strong> later projects. Theelectronic form allows also for easier exchange and shar<strong>in</strong>g <strong>of</strong> resources among translators.Consequently, teams <strong>of</strong> translators work<strong>in</strong>g on large translation projects are equipped with<strong>tools</strong> ensur<strong>in</strong>g greater term<strong>in</strong>ological consistency, and therefore higher quality <strong>of</strong> translation.Also, it is <strong>in</strong>structive to po<strong>in</strong>t out that us<strong>in</strong>g a <strong>term<strong>in</strong>ology</strong> <strong>management</strong> tool isbeneficial, even if other CAT <strong>tools</strong> are <strong>of</strong> little assistance, e.g. when the source text does notconta<strong>in</strong> many repetitions, and there are no parallel texts which can help build a translationmemory. In such cases, the translation memory module may turn out useless, while<strong>term<strong>in</strong>ology</strong> <strong>management</strong> system may be <strong>of</strong> utmost assistance, <strong>of</strong>fer<strong>in</strong>g a significantenhancement <strong>of</strong> term<strong>in</strong>ological consistency (cf. Benis 1998).As it is po<strong>in</strong>ted out by the specialist <strong>of</strong> man-mach<strong>in</strong>e <strong>in</strong>teraction <strong>in</strong> translationprocess:‘[…] an <strong>in</strong>dividual translator cannot carry out the task <strong>of</strong> manag<strong>in</strong>g an entireproject alone <strong>in</strong> a reasonable amount <strong>of</strong> time unless he or she works <strong>in</strong> a team;second, that this team needs to automate as many parts as possible <strong>of</strong> the14


process if it is to provide a quick response to the client; and; f<strong>in</strong>ally, thattranslators need to adapt themselves to this new environment and learn newskills.’ (Rico Pérez 2001)In conclusion, termbanks, and <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong> are <strong>in</strong>dispensable <strong>in</strong> enhanc<strong>in</strong>gthe translator’s ability to transmit a correct message <strong>in</strong> the target language to the recipient.They ensure better quality and consistency and boost the speed <strong>of</strong> translation, reduc<strong>in</strong>g thetime spent on perform<strong>in</strong>g such pre-translation tasks as <strong>term<strong>in</strong>ology</strong> extraction or collectionand validation.5. CONCLUSION:Efficient <strong>term<strong>in</strong>ology</strong> <strong>management</strong> is a prerequisite <strong>of</strong> a good translation service. The<strong>in</strong>creas<strong>in</strong>g volume <strong>of</strong> translation result<strong>in</strong>g from the processes <strong>of</strong> globalization and<strong>in</strong>ternationalization sets new challenges to translators. In order to meet them, all translators,either freelance or corporate, should take advantage <strong>of</strong> the new solutions <strong>in</strong>creas<strong>in</strong>g thespeed <strong>of</strong> translation work while ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g or improv<strong>in</strong>g the quality. Bear<strong>in</strong>g <strong>in</strong> m<strong>in</strong>d theabove-mentioned arguments it seems obvious that efficient <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>in</strong>translation can be implemented only through modern <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong>,tailored to the needs <strong>of</strong> particular work<strong>in</strong>g environments.However, on many occasions translation memory <strong>tools</strong> are perceived as moreproductive and worth <strong>in</strong>vestment than <strong>term<strong>in</strong>ology</strong> <strong>management</strong> <strong>tools</strong>. The cost-benefitratios <strong>in</strong> the case <strong>of</strong> <strong>term<strong>in</strong>ology</strong> managers must therefore be calculated very carefully. Itshould be born <strong>in</strong> m<strong>in</strong>d that the benefits drawn from us<strong>in</strong>g the specialist <strong>term<strong>in</strong>ology</strong><strong>management</strong> s<strong>of</strong>tware become transparent <strong>in</strong> a long-term perspective (Wright:10).15

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!