Towards Identifying the Relationship Between Academic and Professional Criteria for Legal Translation

For many years translation theorists have discussed the degree of translational freedom a legal translator has in rendering the meaning of a legal source text in a translation. Some believe that in order to achieve the communicative purpose, legal translators should focus on readability and bias their translation towards the target language community. Others insist that because of the special nature of legal texts and the sometimes binding force of legal translations, translators should stay as close to the source text as possible, i.e., bias their translation towards the source language community. But what is the relationship between these ‘academic’ observations and the way professional users and producers, i.e., lawyers and translators, think of legal translation? This article examines how actors on the Danish legal translation market view translational manoeuvres that result in a more or less close relationship between a legal source text and its translation, and also the translator’s power to decide what the nature of this relationship should be and how it should manifest itself in the translation.

The answer is a translation based on comparative law analysis, of concepts as well as of formal and syntactic conventions that phrase the legal concepts and the other linguistic and non-linguistic conventions of the ST as faithfully as possible, in a legal language that allows the end user, usually someone of the legal profession, to interpret accurately the legal reality of the ST. (2011,243) Their focus is thus on enabling the receiver of the target text (TT) to understand the content and possibly also the function of the ST (the legal reality), but on the terms of the source language culture rather than the target language culture.
Along the same lines, and concerned with what they refer to as the 'singularity' of the foreign text, Glanert and Legrand urge against the translator "transgressing the text": Perhaps the most challenging enjoinder, therefore, is for the text in translation to deploy itself so as to avoid an assimilation of the foreign text into the host language such that the foreign text's singularity would vanish along the way. As we approach the matter of translatability, the translator's aim must indeed be to preserve something of the difference of the foreign law-text on account of the recognition and of the respect that one owes it, especially as one purports to displace it across languages. (2013,517) Again, other theorists try to balance somewhere in between the two extremes of 'free' and 'faithful', allowing translation creativity but with the proviso that it "must be kept to a 'permissible' (Hammel 2008) or 'relevant' (Hjort-Pedersen 1996) minimum", as Leon Wolff puts it (Wolff 2011, 229).
While insisting on a full and accurate conveyance of the ST meaning, Hammel (2008) focuses on achieving TT readability and discusses whether "a legal translator should render ornate source language legal texts into clear streamlined target-language prose". Drawing on the insights of the plain language movement, he offers a qualified yes to this question (it is permissible in certain cases), the qualification being that the translator must always convey the original's meaning fully and accurately. Hjort-Pedersen (1996), which was based on an empirical study of a corpus of English translations of Danish wills, showed that professional translators quite frequently switch between source and target language orientation in their translation of performative utterances contained in Danish wills. The idea proposed in that article was that relevance theory is able to explain the switches observed in these translations, and that the principle of relevance can be used as a guideline for legal translators in their choice between source and target language orientation when it comes to translating speech acts contained in Danish wills into English.
The argumentation is based i.a. on the following example. In a Danish will, a Danish testator performing the act of creating a will normally produces an utterance of the type shown as (1): (1) Jeg bestemmer herved, at min datter skal arve X which near-literally translates into (2): (2) I hereby decide, that my daughter shall inherit X An English testator performing the same act normally uses an utterance of the type (3): (3) I devise and bequeath X to my daughter So translators will have to make a choice between (2) and (3) as their translation of (1). The choice of (2) means that target language receivers (English lawyers) are presented with an utterance which is unfamiliar to them from similar situations in their own speech community, and (2) will therefore in relevance theoretic terms be more difficult to process than (3). (2) as a translation of (1) will therefore as a starting point violate the principle of relevance. If, however, the translator believes that (2) will add something to the set of assumptions available to the target language receiver that (3) will not add, and that this 'something' is important, then the extra contextual effects produced by (2) will counterbalance this violation. When reading (3) in a translation, target language receivers will automatically activate their subset of assumptions about English inheritance rules. Under these rules testators are, generally speaking, free to dispose of their property by will as they choose. In reading (2) as a translation, however, there is a mismatch between the utterance on the one hand and the will-making context available to the target language receivers on the other. This means that the receivers will have to refer to a will-making context that matches the utterance in question, i.e., Danish inheritance rules, where Danish testators are not free to dispose of their property in the same way. So the extra processing costs involved will be offset by extra contextual effects in relevance theoretic terms. In other utterances, where the extra effort involved in processing a source language-oriented translation does not result in extra contextual effects, target language orientation will be relevant as a translation strategy. Hjort-Pedersen therefore proposed (1996,369) that by switching between source and target language orientation "the translator emphasizes the comparative element which translation between two unique legal systems naturally involves instead of obscuring the differences between the systems".
Wolff is critical of this 'stretch and snap' approach, as he calls it, which in his description allows translational creativity in relation to the ST only to a certain point to avoid that the 'tether' to the ST could be seen as being cut altogether (2011,228). He argues that the focus on text-oriented meaning leads to ignorance of the "contextual meaningfulness of the text as a whole" (2011,230) and goes on to say: "Legal translation theory, in short, needs to break free of its 'stretch and snap' limitations. A free translation that respects the text's contextual foundations should become the new norm." Wolff further claims (2011, 241) that legal translation needs to accept the broader lessons of general translation and legal theory -that perfect (or even adequate) equivalence is a myth (chesterman 1993:3); that meaning and interpretation are not carved in stone (Joseph 1995:14); and that legal systems are not freeze-packed into distinct and definable legal families.
It could be argued, though, that precisely these lessons might lead to the exact opposite conclusion: Language, including legal language, is sometimes underspecified or ambiguous, resulting in word or sentence meaning that is not carved in stone. And to avoid the risk of conveying in their translation a wrong interpretation of underspecified or ambiguous words or sentences in a legal text, the default strategy for legal translators may be to stick fairly closely to the ST and leave it to the reader of the TT to undertake the necessary enrichment or disambiguation of such words or sentences. Also, even though legal systems may not, as Wolff puts it, be freeze-packed into distinct and definable legal families, there is undoubtedly a special relationship between law and language because of the possible normative character of legal texts, and because of legal terms being rooted in specific legal systems. And that might also be an incentive for legal translators to stay close to the ST in order to ensure that target readers are given access to the original ST meaning in such a way that they can decode and enrich the TT not on the basis of their own context but on the basis of the context envisaged by the original author.
Be that as it may, the broader lessons of general translation theory are not unequivocal either, judging by the theoretical prototype of translation hypothesized for general translation by Tirkkonen-condit. Gleaning features from a number of current translation theories, Tirkkonencondit sets up a theory of a prototype of general translation (2011,164). This theoretical prototype reflects a number of assumptions about translation put forward by different scholars, including assumptions that a) Translations presented as translations are assumed to manifest at least some equivalence with the original. b) Translation is based on a source text, and interpretive resemblance is aimed at. c) The translator has the power to decide issues of equivalence, skopos, relevance and sense. As it appears, these assumptions are non-specific as to the degree of equivalence and interpretive resemblance between ST and TT; it is the task of translators to decide what the nature of the relationship should be as they work their way through the translation.
Thus translation theorists may well disagree among themselves as to the goals of legal translation and thus the preferred criteria for and approaches to its performance. But how do professional users and producers of legal translations actually think of legal translation and the options available to the legal translator in the translation process? In line with Jääskeläinen et al. (2011), I believe that "we also need research which will identify the relationship between the 'academic' assessment criteria and the ones prevailing in different sectors of the translation market" (2011,153). Recommendations along somewhat similar lines have been made by I. Schuch, Austrian National Bank, in a European commission report on the usefulness of the clients' views on translation: Translators tend to dutifully render every phrase as written in the source text or as translated before, whereas authors and readers want clarity and readability. Author feedback can make a difference and empower translators to redefine translation quality. Maybe actual clients can be persuaded to share their views. 1

Data
To try to identify the relationship referred to by Jääskeläinen et al. (2011) and the European commission report, I will now proceed to look at viewpoints elicited from Danish lawyers and legal translators on the use of various translation strategies that have a bearing on the nature of the resulting equivalence/interpretive resemblance relationship between a legal ST and its TT. Also, I am interested in examining how the actors on the Danish legal translation market actually view the power of legal translators to decide issues of equivalence, skopos, relevance and sense.
The data consists of opinions voiced by Danish lawyers and legal translators while either performing a specific translation task or while grading different translations of the same legal ST. The data is extracted from three different qualitative studies, one conducted by Faber and Hjort-Pedersen (2013) and the other two by student translators, Fischer (2008) and Nørgård Madsen (2013), in their master theses. 2 The number of informants totals 13 lawyers and 14 legal translators.
In her thesis, Fischer interviewed both lawyers and legal translators about the use of source and/ or target language orientation in a specific translation task that they were asked to perform, i.e., the translation of an English pre-marital contract into Danish, including the reasons for their preferences. As part of her thesis, Nørgård Madsen also focused on source and target language orientation in legal translation, including the use of plain legal language and the Woolf terminology introduced in his 1996 report to make legal texts more accessible to lay readers. In her study, both lawyers and translators were asked to translate an excerpt of the Danish Administration of Justice Act (Retsplejeloven) into English and comment on their strategies and the reasons for them. The Faber and Hjort-Pedersen study focused on the use of explicitation and implicitation in different translations of an excerpt of a Danish law report into English. In this study, lawyers and legal translators were not asked to perform a specific translation task; instead they were asked to grade three different TTs of the same ST. The TTs represented varying degrees of faithfulness to the TT as a result of the use of explicitation and implicitation strategies.
Although the specific purposes of the studies differ, they have one feature in common in that they all attempt to elicit responses from translation market actors -rather than from scholars -to a number of different translational challenges that are all related to the free vs. faithful discussion referred to above. Specifically, the problem areas focused on in the studies include a. the choice between biasing the TT towards the source or target language community, including the handling of culture-bound legal terms; b. the role played by readability and the principles of plain legal language in a TT; and c. the handling of ST ambiguity or non-explicitness in ST phrases, e.g., the possibility of converting ST passive and/or nominal constructions into TT active and/or verbal constructions by adding a syntactic agent.

Theoretical Framework
As a framework for the analysis of viewpoints obtained from the informants I will be drawing on different concepts developed by relevance theorists (Sperber and Wilson 1986;Gutt 1990Gutt , 1991, i.e., the notions of relevance, direct and indirect translation, contextual assumptions and adequate contextual effects, which together provide a principled basis for describing and explaining the multifaceted factors underlying the viewpoints voiced by the informants. For Gutt (1990Gutt ( , 1991, translation has to do with searching for optimal relevance, and the role of the translator is to establish optimal relevance between the producer of the original ST and the reader of its TT as a way to guarantee the success of communication. So how does an utterance, including a translation, achieve relevance? Sperber and Wilson (1986, 125) propose the following definition: a. an assumption is relevant in a context to the extent that its contextual effects in this context are large.
b. an assumption is relevant in a context to the extent that the effort required to process it in this context is small. In Gutt's translational framework, relevance can be achieved either through a 'direct' or an 'indirect' translation. Direct translation corresponds to the idea that complete interpretive resemblance exists between the ST and its TT. Indirect translation involves a looser degree of resemblance. The complete interpretive resemblance relationship in direct translation can be taken to hold only with regard to the original context. To Gutt, this is no extraordinary requirement, but he acknowledges (1990) that it is at variance with a more widely accepted view of translation, where complete interpretive resemblance between the ST and TT is considered doubtful.
The understanding of the two concepts of direct and indirect translation hinges on the different cognitive environments which prevail in the minds of the producer and the receivers. For complete interpretive resemblance to exist between the original ST and its TT, the TT should be interpreted on the basis of the contextual assumptions available to the original ST producer. This means that with a direct translation, a TT will create optimal relevance and adequate contextual effects for TT receivers, though on the condition that they familiarize themselves with the original context in order to derive the meaning intended by the original author.
With an indirect translation, on the other hand, TT receivers can achieve optimal relevance by processing the TT on the basis of the contextual assumptions that are most readily available to them, i.e., the context that they are familiar with from their own speech community. Thus, the concept of 'contextual assumptions' appears to be flexible enough to encompass different types of knowledge about legal systems and language that are either already accessible to different producers and receivers of a legal translation, or that the producers and receivers will have to familiarize themselves with in order to gain access to the meaning of the ST through its TT.
Finally, the concept of 'contextual effects' encompasses the range of different knowledge transfer results that are deemed appropriate in different -in casu legal -translational situations by actors on the Danish legal translation market. In relevance theoretic terms, there are three main kinds of contextual (cognitive) effects, i.e., supporting and thereby strengthening existing assumptions, contradicting and eliminating assumptions, or combining inferentially with them to produce new conclusions. These concepts therefore not only provide a framework for describing different translation choices that will result in a more or less close relation between ST and TT, they also provide a basis for explaining the end result of a particular translation strategy in terms of contextual effects aimed at in the receiver group. This was of particular interest in this study, because I was interested in obtaining a clearer picture not only of what actors on the legal translation market want from a legal translation, but also of the reasons behind their viewpoints.
Obviously, the context and contextual assumptions that are available to receivers of a legal TT may differ very much from the original context anticipated by the author. To take such differences into account, Gutt (2010, 302) operates with a concept he calls 'the notion of congruity' which refers to "the degree of similarity or difference between cognitive environments with regard to the information needed as context for processing a particular utterance (or text)". Therefore, when deciding on choices relating, e.g., to the problem areas outlined above, the translator must consider the congruity of the cognitive environments of both ST and TT receivers in order to try to achieve the desired contextual effects in the receiver group. By way of example, explicitation or omission as translation microstrategies may be triggered by the translator's focus on congruity and a wish to change, strengthen or eliminate existing assumptions in the receiver.
To explore the compatibility of the three sub-assumptions of the Tirkonnen-condit general translation theoretical prototype with the reality of legal translation in Denmark, I will now move on to examine how the viewpoints obtained from the operators on the Danish legal translation market corroborate or conflict with each of the mentioned prototype assumptions and also the reasons underlying the viewpoints expressed, using the relevance theoretical concepts as descriptive and explanatory framework. 2. Translation is based on a source text, and interpretive resemblance is aimed at.
These assumptions are related to the equivalence theories of translation developed by, e.g., Toury (1995) and the relevance theory of translation developed by Gutt (1990Gutt ( , 1991cf. above). The difference between these two theories is that in the relevance theory of translation "it is the translator who believes that he is conveying an interpretation of the source text that resembles the one that the reader of the original might have arrived at. In Toury's equivalence theory in turn those who assume equivalence are primarily the recipients of the target text" (Tirkkonen-condit 2011, 161). For analysis purposes I have chosen to combine these two assumptions, because the viewpoints elicited from the informants switch between these two angles on translation, although the informants do not, of course, use the technical terms.
It appears from the statements obtained that equivalence/interpretive resemblance as a feature of legal translation is indeed assumed by lawyers and translators alike. This may not in itself seem surprising. But the data also shows that the nature of the equivalence/interpretive resemblance relationship preferred by the two groups differs in that the preference of the lawyers can be both one of a close/faithful relationship, in relevance theory terminology tending towards a direct translation, and one of a more free/less faithful relationship, i.e., tending towards an indirect translation. In contrast, the relationship preferred by translators is mainly one of a close/faithful relationship, i.e., tending towards a direct translation. The reasons given by the two groups for their direct translation preferences also differ to some extent.

The Lawyer Informants
The lawyers' preference for a translation tending towards a direct translation is based on two factors, one being a perception of what a translation per se is and is not, the other being whether the translation is meant to be used in a court case, be certified and serve as a legally binding document.

When asked in the Faber and Hjort-Pedersen (2013) study to rank three TTs of a Danish legal ST representing different degrees of closeness between ST and TT, one of the lawyers remarks: 3 (4) Nothing has been added to explain things, and that is the characteristic feature of a good translation, i.e., that it says precisely what the original text said, so to speak […] so you can be absolutely certain that what it [the original text] said is what you read.
Another lawyer was asked in Fischer (2008) to comment on source and target language orientation in legal translation, and remarked:

(5) If I had an employment contract that needed to be translated, I would prefer a very close relationship between the ST and the TT, if it was meant to be used in a court case, or if it was a certified translation. 4 In other situations with both an English and a Danish version of the same contract, I would care less about closeness, because typically you would write 'the English version shall prevail'.
In these cases, the lawyers' translation preference for a close relationship between ST and TT seems independent of the potential context of the receivers (who are not mentioned). The message of the original ST sender is the main focus point, and the lawyers' preferences are thus related to the context envisaged by the original author when producing his/her message.
In other cases, however, where translations are meant to be used as working papers, e.g., for clients and partners in a cooperation, the lawyers tend to prefer a translation which draws more towards an indirect translation. In these cases, the translation strategy preferred is related to the potential context of the receivers. As illustrated by examples (6) and (7) below, the deliberations of the lawyers centre around the nature of the contextual assumptions possibly available to the receiver, the congruity of the legal concepts of the two legal systems involved, and the contextual effects the translation is meant to produce.
In the process of evaluating an English TT containing a number of explicitations of, e.g., syntactic agents that were implied in the Danish ST through the use of passive and nominal constructions (Faber and Hjort-Pedersen 2013), one lawyer comments:

(6) OK, this one is better, but that's because if I were an English lawyer [client] and were unfamiliar with Danish law, then I would get a little more explanation [of what goes on] here.
commenting on the same translation, another lawyer, who also has a degree in translation, explains how she sees her role as a lawyer-translator: (7) I know that this translation represents some sort of interpretation, but I do that as well myself. Sometimes I make translations into French […] and if I know that the receiver will not understand it, then I add explanations.
The nature of the contextual assumptions possibly available to the receiver, the notion of congruity and the contextual effects the translation is meant to produce are indeed very important aspects for the lawyers. But the issue of time spent on a case and the financial costs involved in producing translations may remove this importance, at least to some extent:

The Translator Informants
Turning to the translators, their preference for direct translation is based on two factors. One factor resembles that of the lawyer group, i.e., whether the translation is to be certified and serve as a legally binding document. Moreover, one translator knows from experience that a lawyer who commissions a legal translation may change the purpose of the translation from one day to the next, and that is a situation which is hard to handle. This is illustrated in examples (9) and (10) below: (9) They cannot always understand that if I made a translation that was not originally meant to be certified, then I will have to make it sort of 'better' [if the need for certification subsequently arises]. But I cannot really explain that [to the lawyer commissioner] because then his reaction would be 'oh really, so the first one was not good enough'. […]. Therefore, I believe that the translation should always be of a type that can be certified. (10) We as translators are sometimes told that we are too faithful to the ST -but then again, you never know whether at the end of the day your translation suddenly needs to be certified.
The second factor mentioned by the translators -in contrast to the lawyers -has to do with the status of the ST as an authoritative document in itself. In Nørgård (2013), three translators talking about the translation into English of an excerpt of the Danish Administration of Justice Act remark: (11) The legislators wanted the wording to be like this, so you need to be careful not to change it […] (12) The text type determines the strategy. [If it's] legislation that needs to be translated, then I would personally avoid interpreting too much, e.g., by inserting a subject that is not linguistically realized in the ST. […] -the more legal it gets, the closer you naturally stick to the ST, I believe. I don't mean to say that it shouldn't be fluent […] in the target language, but the ST is the important thing. This is not a situation where you try to produce beautiful linguistic constructions. (13) I wouldn't want to change the layout radically [...] because we are dealing with an authoritative document in Danish, and therefore it's not OK to [...] present the text in a completely different way.
The above examples show that the direct translation preference of the translators in these situations -as with the lawyers -is independent of the potential context of the receivers; rather, it is related to the context envisaged by the original author.
Unlike the lawyers, however, the direct translation preference of the translators is also in some situations related to the potential context of the receivers. In such situations, the translators take into consideration the nature of the contextual assumptions that the receivers are expected to have, the notion of congruity and especially the risk of an indirect translation strategy creating undesired contextual effects by prompting the receiver to produce wrong new assumptions and conclusions.
In Fisher (2008), a translator commenting on the use of source and target language orientation in her Danish translation of an English pre-marital contract says:

(14) The aim of my translation is that a Dane without prior knowledge of English law understands the content of the contract. That means that my translation must be of a kind that signals that we are dealing with two different legal systems. (15) […] so I wouldn't bias it too much towards the target language [because of the risk of]
making the reader think that the legal systems are the same.
Another translator commenting in Nørgård Madsen (2013) on her English translation of an excerpt of the Danish Administration of Justice Act makes rather similar observations: (16) The older I get, and the more experienced I become, the more I realize the danger of applying a target language orientation, because you risk making the receivers believe that […] the legal content of a term is similar to what they know from their own speech community.

Assumption 3
3. The translator has the power to decide issues of equivalence, skopos, relevance and sense

The Lawyer Group
Even though the lawyers prefer indirect translations and thus a more free relationship between ST and TT in certain communication situations, they are not always altogether comfortable with the translator deciding the way in which a more free relationship should manifest itself in the TT.
One lawyer in the Faber and Hjort-Pedersen (2013) study argues: Another lawyer, commenting in the same study on a TT containing a number of explicitations, mentions the risk of a more free relationship between ST and TT resulting in doubts in the receiver group as to who is speaking, the translator or the ST.

Assumption 3
3. The translator has the power to decide issues of equivalence, skopos, relevance and sense

The Lawyer Group
Even though the lawyers prefer indirect translations and thus a more free relationship between ST and TT in certain communication situations, they are not always altogether comfortable with the translator deciding the way in which a more free relationship should manifest itself in the TT.
One lawyer in the Faber and Hjort-Pedersen (2013) study argues: So, as it appears, these lawyers are not altogether at ease with the translator having the power to decide which extra information is relevant to achieve an optimal resemblance between ST and TT.

The Translator Group
Even though the preferred translation strategy of the translators generally draws towards the direct translation type, some of the translators admit that indirect translation and thus a more free relationship between ST and TT rather appeals to them because they consider such translations easier to read and understand. But, as with the lawyer group, they do not seem to believe that the translator really has the power to decide the way in which a more free relationship between a ST and a TT should manifest itself. One of the reasons rests on a perception of the lawyer commissioner as having more authority than the translators themselves in a particular translation situation:  (9) and (10) above, is that translations made from one skopos perspective may prove inadequate if a sudden change of skopos is forced upon the translators by a lawyer commissioner because a translation made for informative purposes suddenly needs to be certified. The translators feel unable to cater for this possibility -to them each translation is a unique textual unit that cannot be recycled and reused unaltered in other situations.
Moreover, the translators are also aware that legal ST vagueness/underspecification may be deliberate on the part of the drafters of the ST. They may have no way of ascertaining whether that is in fact the case in a particular legal ST, and therefore they feel that they do not have the power to decide whether to opt for some sort of specification and thus a more free relationship between ST and TT. One translator in the Faber and Hjort-Pedersen (2013) study argues along the following lines: (23) Hmmm, I wonder whether this was intended, I think that can be really difficult [to determine], and then I try to reproduce the ambiguity or vagueness.
In their own perception, the limitation in the translators' power to decide is in these cases defined with regard to the context envisaged by the original author.
But it may also in some cases be defined with regard to the potential context of the receivers, because translators are afraid of their translation not producing adequate and therefore relevant contextual effects, i.e., it does not add to existing assumptions because it tells a lawyer receiver something that he already knows. commenting again on the translations containing a number of explicitations in the Faber and Hjort-Pedersen (2013) study, one of the translators argues: (24) Because the people who read this type of text know very well what this is about.

Conclusion
As outlined above, legal translation theory has for many years sought to influence the practice of legal translators by advocating either a free or a faithful relationship between a legal ST and its TT, or maybe even something in between. However, very little effort has been made to investigate how 'academic' recommendations actually correlate with the preferences of professional actors on the legal translation market, what the reasons for their preferences are and also who has the power to decide what the nature of the ST/TT relationship should be and how it should manifest itself in a legal translation.
Drawing on the descriptive framework of relevance theory, this article provides insights into the relative socio-cultural and professional status of legal professionals and translators, and it suggests that the preferences of the professional actors on the legal translation market are not really clearcut. The lawyer informants seem to prefer translations of both the direct and the indirect type. They prefer a translation of the direct type in situations where a translation is to be used in a court case or is to be certified. In these cases, readability for the receiver is not in focus, but the ST and the contextual assumptions envisaged by the original author is the important thing, and it is up to the receivers to familiarize themselves with this context to be able to achieve relevance and adequate contextual effects. The extra processing effort required on the part of the receiver in these cases is deemed relevant because extra contextual effects are deemed large and will counterbalance this effort. In other cases, where translations are to be used in a client relationship, the lawyers are very much in favour of translations of the indirect type, where the information needs of the receiver are catered for in the translation, subject however to the question of time spent and costs involved. In these cases, indirect translation is deemed relevant because the effort required on the part of the receiver to achieve adequate contextual effects will then be small.
For the translators, on the other hand, it is not a case of 'both-and'. They agree with the lawyers on a direct translation being preferable in situations where the translation is to be used in a court case or is to be certified, but their preference for this type of translation is not restricted to these situations. Part of their reasoning mirrors that of the lawyers in that readability for the receiver is not in focus. Rather, the context envisaged by the original author is the important aspect. But another part of their reasoning differs from that of the lawyers. This part of their reasoning is related to the context of the receivers because the translators are focused on the risk of a target language orientation creating undesired and irrelevant contextual effects by prompting the receiver to produce wrong new assumptions and conclusions.
As for the translator's power to decide, the data seems to suggest that both groups of informants are not altogether at ease with translators deciding to read their own interpretation of a legal ST into their TT, e.g., by adding information that is not linguistically realized in the ST, resulting in a translation drawing towards an indirect kind. For the translators, doing so would in a sense mean that they were 'competing', either with the original sender of an authoritative legal document (who opted for a specific wording or layout) or with a lawyer receiver (who is the legal expert and does not need to be told things). The lawyers, on the other hand, are afraid of being 'deceived' by an indirect translation, because they have no way of knowing 'who is speaking when', the translator or the translation. It therefore seems that the translator's power to decide on freeing the TT from the ST may be more restricted in legal translation than in other types of translation.
I am of course aware that the data constituting the basis for this article is limited and relates to the Danish legal translation market only. The data seems to suggest that there is no one 'best practice' model for legal translation in Denmark: it depends who you ask and on the circumstances. However, it may be argued that this conclusion might well extend to legal translation in other countries also, because the reasons given by the informants are general in nature in that they do not relate to specific legal cultures or areas of law.
There is one aspect that would indeed be interesting to investigate also, i.e., the translation preferences of producers and users of EU translations that are to function in contexts rather different from the ones focused on in this article, because with EU translations there is no source and target culture in the usual sense. However, that topic is not something that has been touched upon in the data contributed by the informants constituting the empirical basis for this article, although it would certainly be a highly relevant focus of future research projects.
Still, I hope that this article provides some explanations of the relationship between academic criteria for legal translation and the ones prevailing in different sectors of the legal translation market.