ORGANIZATION OF BACKGROUND KNOWLEDGE STRUCTURES IN LEGAL LANGUAGE AND RELATED TRANSLATION PROBLEMS

: Th e paper examines the organization of background knowledge structures in legal language and related incongruities of legal terms. Th e cognitive linguistics methodology, in particular its fi ndings on the nature of meaning, is applied. Terms serve as prompts with a semantic potential to activate various levels of knowledge structures, such as domains, scripts, scenarios, cognitive models and frames. In most cases organization of knowledge will diff er in the SL and TL. Th e fi nal part analyses translation strategies and techniques in terms of their potential to activate relevant knowledge.


Introduction
Translation is frequently regarded as an act of communication, and legal translation is defi ned by Šarčević as a special type of act of communication which takes place "in the mechanism of the law" (2000).Our conventional model of communication is refl ected in the conduit metaphor discussed by Reddy.People talk and think about linguistic communication as about the sending and receiving of parcels fi lled with meaning.Ideas, thoughts, emotions are taken out of the mind and put into words by the speaker; next they are sent along a conduit to the receiver, who unpacks the ideas from 1 A researcher and translator trainer in the Department of Translation Studies and Intercultural Communication, Institute of English, University of Gdańsk, Poland, as well as freelance and sworn translator.She holds a Ph.D. in Linguistics (University of Gdańsk), M.A. in English Studies/Translation (Jagiellonian University), and a Diploma in an Introduction to English Law and the Law of the EU (University of Cambridge).
words (Ungerer and Schmid 1996:119).Deeply ingrained as it may be, this model is misleading in a number of ways.It suggests that it is possible to pack meaning into a word and fi nd another packaging for that content in another language (translation comes from Latin transferre = to transfer; Oxford English Dictionary).It has been experientially proved that human communication is much more complex due to the nature of concepts.Th e aim of this paper is to analyse the organization of background knowledge structures behind legal concepts and related translation problems.

Semantics of legal terms and concepts
Terms, as argued by Sager, are 'depositories of knowledge' and units with specifi c reference in that they "refer to discrete conceptual entities, properties, activities or relations which constitute the knowledge space of a particular subject fi eld " (1998:261).Although the discreteness of concepts may be questioned, the claim that concepts are embedded in complex knowledge structures is in line with the approach to semantics proposed by cognitive linguistics (CL), the mainstream branch within functional linguistics.CL views language as a direct refl ection of cognitive processes and an integral part of human cognition and understanding of the world.In Langacker's 'subjectivist' theory of semantics, meaning is understood as a dynamic process involving conceptualization (mental experience) rather than a static bundle of features (1988:50).As argued by Evans et al., during conceptualization "linguistic units serve as prompts for an array of conceptual operations and the recruitment of background knowledge" (2006:160); therefore, they have a certain semantic potential (Evans 2006:493).Th is claim also extends to terms, which function as prompts or points of access to vast knowledge structures rather than as containers for knowledge.Th e connection between a term and its knowledge structures may be seen as a mental routine.Sometimes people know only 'linguistic labels' but have not built the mental path which activates relevant knowledge: you may know the term constructive trust but may have diffi culties in explaining what it means.
Th is dynamic approach is based on Haiman's claim that meaning resembles an encyclopaedia rather than a dictionary.Th erefore, to characterize a concept it is necessary to refer to other cognitive domains presupposed and incorporated by it (Langacker 1988:53).To understand a tort of negligence, the following domains should be activated: tort law, court, duty of care and breach of duty of care, civil liability, loss and remedies.Th e tort of negligence also has a semantic potential for activating the domains of the Caparo test, defences, causation, solatium, etc. Th e domains overlap and are arranged hierarchically in terms of relevance and salience.Depending on a usage event, some domains are activated in the foreground, some are activated in the background, while others are not activated at all.Th e foreground is the focal point and is referred to by Langacker as a profi le, the background serving as a base (1988:60).Th e term base emphasizes "the way in which background knowledge 'supports' the concept" (Clausner & Croft 1999:5) by providing a context.
All the domains activated by a given concept are called its matrix (Langacker 1988:56).Since meaning is a mental process, it invites idiosyncrasy: the domains activated by conceptualizers will diff er depending on their experience and knowledge.Th e defendant may activate the domain of defences with more detail, focusing on the voluntary assumption of risk, contributory negligence or illegality while the claimant may focus on the domains of loss and remedies.A judge is expected to have a more structured knowledge and activate more domains than a lay magistrate from the magistrates' court or a working-class tortfeasor aged 19.
For this reason it may be diffi cult to specify how many domains have to be evoked to understand a legal term.According to Shelov's degrees of terminologicality, the more information is required, the more terminological a lexeme is (qtd. in Th elen 2002:196).In the case of legal concepts, the supraindividual semantic potential of terms is specifi ed in legislation and case law.A substantial part of this knowledge is expected to be internalised and intersubjectively shared by members of the legal profession.
As emphasised by Clausner and Croft , the structure of domains "is more than a list of experientially associated concepts" (1999:2).Knowledge is arranged not only in terms of relevance and salience; concepts form complex interrelated networks and such vertical and horizontal interrelations are part of their meaning.A concept is understood fully when the conceptualizer knows its exact place in the network (cf.Sager 1998:260).First of all, concepts are organised in terms of their level of specifi city along taxonomic vertical hierarchies (Langacker 1988:64).For example, a legal person is more schematic than a company; hence, the image it evokes is less rich in detail.Secondly, legal concepts are frequently organised along horizontal causal scripts/scenarios.Kjśr argues that the horizontal organization has a form of complex IF-THEN rules that prestructure the expert's knowledge and reasoning patterns.A legal concept connects legal conditions with legal eff ects; the connecting concept may be regarded as "a reduced representation of legal rules " (2000:146).A similar view is expressed by Gizbert-Studnicki, who sees legal terms as shortcuts that connect a certain set of facts with a certain set of legal consequences.He notes that the sets of facts and consequences are unlikely to be identical in two legal systems and goes as far as to suggest that the connecting concepts known as 'legal institutions' are proper names and as such are untranslatable (2001:52).Th ese assumptions can be illustrated by tort of negligence, which evokes the following scenario: if the defendant owes a duty of care to the claimant and breaches it, the claimant suff ers a loss, the loss is not time-barred and the tortfeasor is not able to raise any defences, then the claimant may bring a civil action and claim damages.Th is generic scenario forms part of the meaning of tort of negligence and is fi lled with details in a usage event.
Besides causal relations, concepts are embedded in various cultural models that organise a given fi eld, which is clearly visible from the crosslinguistic perspective.For example, Board of Directors evokes the onetier corporate governance model, where the board has both supervisory and management functions.By contrast, the Polish term Zarząd, which is provided in 5 out of 6 legal and business dictionaries as an equivalent of the Board of Directors, automatically activates the two-tier corporate governance model, which clearly separates management functions (Zarząd) from supervisory functions (Rada Nadzorcza).Th e incongruity between Board of Directors and Zarząd results from their entrenchment in diff erent cultural models.
It may be supposed that concepts, scripts and scenarios are embedded in a number of overlapping, larger structures called frames.A frame is defi ned by Fillmore as a system of interrelated concepts that form a coherent script-like structure (1982:117).Ungerer and Schmid see frames as a type of cognitive models which include scenarios, domains, interactive networks and scripts, and claim it represents "a cognitive, basically psychological view of the stored knowledge about a certain fi eld" (1996:211).Th e most important frame is the metaframe formed by the legal system itself.It provides organising principles, rules of legal reasoning, approaches to statutory or contractual interpretation.For example, it is more frequent in continental countries to apply the teleological approach to statutory interpretation, while English courts prefer the literal approach (cf. McLeod 2005:328).Another illustration may be found in UK and US contract law and its parol evidence rule.Th e rule prevents any modifi cation of contracts with evidence that was not included therein.By contrast, according to Article 65(2) of the Polish Civil Code, intentions of the parties should be taken into account: "In respect of contracts, one should determine the congruent intention of the parties and the purpose of the contract rather than rely on its exact wording." Th erefore, the approach to interpretation may limit or widen the scope of knowledge activated during meaning construction.
Th e macroframe consists of a number of microframes which function as a narrow context.Some frames may be shared by legal systems, especially those that stem from the same legal traditions, such as the continental legal systems.Polish frames will show more similarity to French or German frames than to English ones.Owing to the harmonization of law in the EU, the incongruity of frames between the continental systems and the UK (common law) system will decrease in time.
Th e above analysis provides a deeper insight into the nature of incongruity of legal terms between languages and legal systems.Th e incongruity applies not only to the boundaries of concepts underlying terms but, above all, is connected with the complex organization of knowledge structures in the SL and the TL.Concepts are embedded in diff erent macro-and microframes, cognitive models, scripts, scenarios and domains, and it is very unlikely that these structures will be organised in the same way in two languages/legal systems.In most cases terms will not have the same semantic potential in the SL and the TL.

Textual conventions
Conventions concerning the amount of knowledge to be incorporated into a legal text may differ in the SL and TL country.Hill and King's comparative research of US and German contracts has shown that the former are significantly longer than the latter.This is because less statutory contract law is applicable to transactions in the US.The US legal system is not uniform as it has fifty state statutes and the common law.Some states, such as Louisiana and California and several Western states, are more code-oriented, while others regulate only certain types of contracts by statute (2004:921).Hill and King's findings may be extended to Polish contracts, which in fact are similar to German ones.Poland is a civil law system with only one jurisdiction; contracts are governed by the Civil Code and hence show a higher degree of intertextuality.Additionally, some contract law has a form of definitions and elaboration of typical legal concepts.As emphasised by Hill and King with reference to US contracts, "the parties typically would not have available either a common-law or statutory definition that could be readily incorporated" (2004:913); hence, contracts contain a large amount of knowledge, detailed definitions and litanies of synonyms.US lawyers attempt to draft very cautious, self-contained contracts since they may be litigated in various states.Having regard to the foregoing, it may be supposed that a verbose US contract translated into Polish and thereby uprooted from its legal context will be easier to understand than a concise Polish contract translated into English.
It is worth emphasising that a source text contains as many prompts and activates as much knowledge as necessary for SL recipients to understand it properly, which is consistent with Grice's Maxim of Quantity.Diffi cult as it may seem, the translator's task is to ensure that TL prompts are adequate for TL recipients in that they create the same legal eff ect as SL prompts.Th is may pose problems in legal translation since, in general, translators are not expected to make explicit information that is implicit in the source text.

Knowledge and translators
With regard to LSP translators, who frequently work under time pressure, they should have suffi cient knowledge to limit the time required for research.Th e translator's knowledge is one of the factors that aff ect translation quality and productivity.Th is requirement is emphasised not only by theorists (e.g."translation as knowledge-based activity", Wilss 1994), but also by translator educators and the translation industry.For example, the programme of European Master's in Translation prepared by the Directorate-General for Translation (EC) recommends allocating at least 50% of the total credits to practical translation classes and courses fostering knowledge of special fi elds and their languages.With regard to the translation industry, the translator's knowledge of the fi eld and expertise are regarded as an industry standard and are frequently provided for in contracts that regulate the relationship between freelance translators and translation agencies.Additionally, some institutions set very high requirements concerning the translator's formal training in the special fi eld.Th is can be illustrated by the European Court of Justice, which employs only lawyer-linguists as translators.
In respect of legal translation, the translator should know both the SL and TL legal macroframes: the more specifi c and detailed the frames, the higher the quality of translation and the lower the risk of mistranslation.As emphasised by Wilss, if the translator has little expertise, s/he will be involved in local processing of information without the wider picture rather than more global and effi cient processing (1994:41).In particular, the translator should be able to distinguish a term from a mere word and select the right equivalent from a dictionary entry.My analysis of English equivalents of Polish company types shows that in 5 dictionaries there are as many as 16 equivalents of spółka akcyjna, some of which are incorrect (Biel 2006).Th e overall quality of bilingual legal dictionaries published in Poland leaves much to be desired, which poses constant challenges to translators and increases the time they need to fi nd the most suitable equivalent.
In addition to the knowledge of the fi eld, another crucial component of translation competence is the knowledge of the target audience's expectations.As noted by Séguinot, "a successful translator is conscious of the receptivity and reactions of the target audience.Some of that recognition may be stored in the form of knowledge, but each new client and every new assignment brings with it the potential for interaction and novelty" (2000:96).Generally, the translator is expected to fi ll in the knowledge gaps of the target audience; however, this issue raises some controversy in legal translation.Th e infl uential Polish Sworn Translator Code claims that the translator is entitled to assume that the recipient is aware of incongruity of legal systems; hence, the translator does not have to provide any additional explanations or defi nitions (Kierzkowska 2005:87-88).On the other hand, in TL-oriented approaches, such as Šarčević's receiveroriented approach, it is argued that translators should "compensate for conceptual incongruity whenever possible" to ensure that the SL text and the translation have the same legal eff ect (2000).Th is, however, requires the translator to project how SL and TL texts will be received, and hence to be more actively involved.

Translation strategies and their potential to bridge knowledge gaps
As already noted in section 3, in most cases terms will not have identical semantic potential in the SL and the TL.Furthermore, the translator may well be able to replace the SL concept with a relatively similar concept from the TL legal system, but the TL concept will not be able to evoke the same knowledge structures.Th e scope and depth of knowledge evoked will depend on the recipients, who may range from experts to laypersons.As noted by Schäff ner, the recipient's prior knowledge has a crucial impact on his/her ability to understand a text (1993).Th ere are several degrees of terminological incongruity, ranging from identical concepts (very rare) or near equivalence to conceptual voids without equivalents in the TL.Th e techniques of dealing with incongruous concepts may be placed along the continuum between two extremes: domesticating and foreignizing strategies.As noted by Venuti, the debate between domesticating and foreignizing is long-standing in translation practice.Domesticating involves assimilation to the TL culture and is intended to ensure immediate comprehension; hence, it is also referred to as the TL-oriented strategy.By contrast, foreignizing "seeks to evoke a sense of the foreign" by "sending the reader abroad"; as a result, it may pose a risk of incomprehension (2001:240-4).It is also known as the SL-oriented strategy.
Chesterman proposes to reserve the term strategy for macro-level problem-solving, a cognitive plan, e.g."the initial choice of source or target orientation, decisions about foreignizing or domesticating", while the term technique should be used for "routine, micro-level, textual procedures" (2005:26).Th is distinction will be followed in the subsequent section, which discusses major translation techniques in terms of their potential to activate relevant knowledge structures.Th e micro-level techniques of legal translation were adopted with some modifi cation aft er Weston (1991:19-34) and Harvey (2000).Th e most important ones include: transcription, literal equivalent, descriptive equivalent and functional equivalent.Th ey may be placed along the foreignizing-domesticating continuum as shown in Chart 1. SL-and TL-orientation in legal translation are based on Kierzkowska's distinction between near and far recipients.Th e former have a relatively good knowledge of the SL culture, while the latter do not know it and have little motivation to learn it.In the fi rst case SL-oriented equivalents of legal terms should be used to emphasize diff erences, while far recipients require TL-oriented equivalents to capitalize on similarities (2002:88-89).
Th e fi rst technique is transcription (borrowing), including naturalization (adaptation of spelling).It is the most foreignizing strategy, the use of which is usually motivated by a large incongruity or untranslatability of TL concepts.As Weston notes, "inevitably ... there will be a number of SL expressions which defy translation in the strict, narrow sense because nothing truly comparable to the corresponding concept exists in the TL culture and a literal translation makes no sense " (1991:26).Prominent examples include common law, equity and trust.Th e advantage of using a transcription is that it provides a clear and accurate reference to the external source frame.As emphasised by Šarčević, it specifi es "the law according to which national terms and institutions are to be interpreted", which is very useful in translation of international conventions "because national courts have no other choice but to apply the foreign concept" (2000).From the translator's point of view, a borrowing is a 'safe' equivalent as it allows him or her to avoid liability for inaccuracy.On the other hand, accuracy is achieved at the expense of comprehension.Some researchers emphasise that this technique "admits defeat" (Weston 1991:26).Th e meaning is opaque and has low analysability; it does not capitalise on the TL knowledge.In consequence, its understanding requires more mental eff ort and a good working knowledge of the SL legal system.Th is technique is more appropriate for translation from well-known languages, e.g. from English to Polish rather than from Polish to English.Given the low popularity of Polish as a foreign language and few English-language publications on Polish law available in the UK and the US, transcription of Polish legal terms in the English texts would be impracticable.Th ese barriers are lower in translation from German or French.
Another technique, literal equivalence, may be regarded as a special type of borrowing.It is also known as formal equivalence, word-for-word translation, calque or loan translation.As noted by Weston, the acceptability of literal equivalents depends on their type.Some do not correspond to any TL concept (neologisms) but are suffi ciently transparent in meaning; in some cases, it is possible that a literal equivalent will also be a functional equivalent.Literal equivalents are not acceptable when they are false friends (refer to a diff erent TL concept) or are virtually meaningless (1991:25).For example, the term limited partnership is calqued in some dictionaries as spółka z ograniczoną odpowiedzialnością, which is misleading since it evokes the concept of a Polish limited company.Th e danger connected with the literal equivalent is that it will neither refer the recipient to the right SL frames nor, as with most neologisms, be connected with knowledge structures in the TL.Th e latter may be exemplifi ed by the Roman-law term użytkowanie wieczyste, which is unknown in the common law system.Th e standard equivalent in legal dictionaries and translation practice is perpetual usufruct.Obscure to most English lawyers, this term does not evoke the information that użytkowanie wieczyste applies to land owned by the state or local government and leased for 99 years or less but at least 40 years (Articles 232 and 236 of the Polish Civil Code).
Instead of perpetual usufruct, the translator may use a descriptive equivalent (also known as a gloss or a paraphrase), such as 99-year or less lease of land owned by the state or local government or a long-term leasehold.Th is technique is more TL-oriented than the previous ones as it takes into account the recipient's knowledge gaps.It is based on explicitation, i.e. making explicit in the TT what may be implicit in the ST.It is worth noting that explicitation is not so infrequent in translation practice: it is regarded as one of the translation universals (Laviosa-Braithwaite 2001:289).Th e descriptive equivalent may provide more (but not complete) information than the literal equivalent and is certainly more comprehensible.Th e major disadvantage of descriptive equivalent is its length.Since a term functions as a shortcut and may be repeated frequently in a text, one of its main properties should be brevity of form.Th e longer the equivalent, the more inconvenient it is.
A descriptive equivalent is frequently based on a legal term known in the TL but which undergoes some modifi cation to signal the diff erence.It is argued that foreign-sounding equivalents make recipients realize the incongruity of terms and refer them to the proper legal system (cf.comprehensibility.In fact, legal translation may be perceived as a hybrid where the SL text is accessed through TL knowledge structures.What requires further research is the nature of mapping from TL concepts onto SL concepts as well as the role of TL knowledge structures in understanding a legal translation.