Many translation companies offer a wide spectrum of language services, which often includes specialised legal translations. However, in what ways is a legal translation different from other specialised translations? What do you need to keep in mind when you entrust a translation company with the translation of your legal documents?
What is a legal translation?
Put simply, it is a specialised translation of a legal document. Nevertheless, this encompasses even more types of texts than you would first expect. A specialised translation in this field can include much more than contracts, court rulings, legal options, GTCs or statutes. Employment contracts, compliance guidelines and patient admission information for a hospital stay – to name just a few – are all documents that contain legal references and regulations.
More than just legalese
Examples such as employment contracts or patient information forms also reveal why legal translators need to be familiar with more than just the legal side of things. Translating employment contracts requires HR expertise as well, and a patient information form can only be translated accurately if the medical terms it contains are understood correctly. Whether it is a contract, court ruling, statutes or compliance guidelines, the main task is organising non-legal elements within a legal structure. Therefore, specialised legal translators must be capable of familiarising themselves with new industries and terminology as rapidly as possible.
A legal jurisdiction and its language(s)
In general, there are two possible scenarios for legal translations:
1. One legal jurisdiction with multiple languages
In this scenario, the legal jurisdiction in question has several languages with equal standing. Examples of this would be Switzerland, Belgium, the EU and the UN. However, the legal jurisdictions of both the source and target texts are the same. If a legal jurisdiction has more than one language, the languages nevertheless all refer to the same legal entity.
As a result, on the official European Union websites you can assume that ‘engagement’ in family law means the same as ‘Verlobung’ in German, ‘fiançailles’ in French and ‘fidanzamento’ in Italian. When dealing with official languages, translators must ensure that the words are actually used in the respective language version. Despite the apparent similarity, translating ‘engagement’ as ‘promessa nuziale’ in Italian would be wrong because the official EU terminology database, the IATE, only provides ‘fidanzamento’. Using another term could lead to confusion and misunderstandings.
This means that research must be done to determine which terms are officially approved and which are used more randomly. A practised feel for language is required in this case, supported by the corresponding terminology databases, to make a definitive decision. In day-to-day business, this would be relevant for court rulings, for example. A translator must decide on a case-by-case basis which formulation originates from the deciding authority and which from the respective legislation.
Regional or minority languages recognised by the European Charter are an exception to the rule because the sources of law are often not available in the respective language. However, there exists a ‘right to use’ of these languages in legal processes. Translations in this area often demand a high degree of creativity, which is based, among other things, on a precise knowledge of the structures and morphological idiosyncrasies of the languages in question as well as the legal concepts.
2. Multiple legal jurisdictions with multiple languages
The second scenario usually deals with cross-border issues in which a different legal jurisdiction applies for the source text and for the target text. If the languages of the source and target texts belong to different legal jurisdictions, this will also have a direct effect on the translation process. In such cases, identifying the terminology provided is less of a problem than searching for comparative terms.
Legal institutions with different legal jurisdictions do not generally correlate with one another, which is why it is an undisputed fact that legal translations are impossible to complete without conducting a legal comparison. This challenge can also occur within the same country, such as the special case of Canada. The legal jurisdiction in the English-speaking provinces is based on common law, while private law in the French-speaking province of Quebec is based on French-heritage civil law.
The function of legal language – and the consequences thereof
Language has a particular function in law. And this has an effect on translations as well. In earth and engineering sciences, language is used to describe ‘what is’. In the humanities – which includes law – things come into being only once they have been described. In other words, language creates things. Unlike in medicine or technology, in the humanities you cannot simply point to something and say ‘That’s what I mean’. Language must always be used to explain what you mean
As a result, even without context you would be able to verify if the German ‘Tulpe’ is referring to the same thing as ‘tulip’, but you would not be able to verify if ‘Sicherheit’ is referring to physical ‘safety’ or online ‘security’, for example. This means that more research is usually required for legal translations than for scientific or engineering translations.
Often similar, never the same: Legal terms in different languages
The fact that legal terms are purely intellectual constructs has concrete consequences. The word ‘robo’ in a Spanish court ruling can never refer to a ‘theft’ in English, but rather only to the specific crime defined in the Spanish ‘Código Penal’ criminal code because ‘robo’ in Spain and ‘theft’ in the UK are two different concepts. In Spain, violent acts enacted against a person or property are also classified as ‘robos’. However, the Theft Act 1968 does not refer to violence at all but rather solely to ‘the dishonest appropriation of property’.
Section 8(1) of the 1968 Act, on the other hand, defines ‘robbery’ as an act of theft with force used against a person. So even though ‘robo’ is often translated as both ‘theft’ and ‘robbery’ in multilingual dictionaries, this means that online dictionaries are of no help either without comparative legal expertise. A simple solution to this dilemma would be to use the Spanish word ‘robo’ in English and add ‘according to Art. 237 et seq. of the Spanish Código Penal criminal code’ when you use ‘robo’ for the first time.
If the text contains no further elaborations on the type of crime committed, the translator would also have to take that into consideration in the translation. The ultimate aim of the translation is to ensure that its readers will understand it in the same way that the original’s target audience understands the source text. Concrete solutions could include adding an explanatory comment when the word appears the first time in the text, and then introducing and using a shorter form whenever it is mentioned again in the rest of the translation. In the case of a theft with property damage, it could look something like this: ‘Theft with violence against property according to Art. 237 of the Spanish Código Penal criminal code (hereafter: aggravated theft)’.
Legal texts and machine translations
Ever since the qualitative leap achieved by the implementation of neural machine translation systems, the resulting translations read more smoothly and fluently than ever before. From a content perspective as well, you might even find yourself thinking that the machine ‘understood’ what was meant. The natural flow of the output can often lead to the impression that the content is also completely correct.
However, this is problematic when it comes to legal translations in particular because the slightest nuances are what matter in such texts – and these are often not recognised by computer programmes. This is frequently the case for double negatives. Nested sentences with multiple subordinate clauses also cause serious problems for these tools on a regular basis, meaning that the likelihood is very great that the resulting translation contains the complete opposite of what the source text says.
The issues mentioned above apply especially to machine translations offered online with programmes that have been trained using a large number of usually heterogeneous texts. The question of data security also arises in such situations because these online machine translations save the respective input on their servers at least temporarily. Programmes that are trained specifically for a certain text type perform somewhat better.
In addition, if the source text is written with an eye towards its future translatability, this can also improve the programme’s output. Nevertheless, both approaches require a high degree of effort that does not always pay off in the end. So, although a machine translation can provide a quick solution, at this point in time it is either unreliable or expensive – or both.
Professional specialised translations in the legal field require not only linguistic knowledge and excellent research skills but also a high level of familiarity with the law and comparative legal expertise. Legal translations themselves are often the result of legal deliberations, so when you are choosing the right translation agency, special attention should be paid to the translators’ legal expertise. Machine translations are to be used only with extreme caution in this industry. Firstly, because there are significant data protection concerns to consider, and secondly because machines that have not been trained in the specialised area are often unable to cope with the complexity of technical legal jargon.