Dilemmas of style when translating legislation

Many dilemmas of style arise when translating legislation into English. This post looks at the most common ones and solutions to them.

By legislation I mean any laws or rules set down by a governing body, be that of a country, company or university. So everything from a criminal code to a health and safety policy.

Recast or follow source structure?

Should you cut and join sentences and shuffle parts of them when translating legislation?

Doing so can feel like you’re overstepping the line. But it depends on your goal. If your aim is comprehension, why not do what you can to make the text readable and understandable?

Special cases do exist. But these may serve as evidence for why we should recast and shuffle.

For instance, the different language versions of EU legislation must have the same number of sentences as part of a policy to make all versions equivalent. But the EU is going for something other than translation here. It wants “proof of linguistic equality” and is willing to sacrifice communication for this end (Pauino, 2016, p. 5). This has even resulted in the appearance of a new type of legalese: “union legalese” (Trosborg, 1997, p. 152).

The EU’s policy of linguistic equality does not put communication first as this is not it’s prime concern. Therefore, if you want to aid understanding in your translation, don’t worry if you’re not following the EU’s convention of mirroring source structure. Translate legislation as you would any legal document. Be accurate but help the readers as much as possible and recast as necessary.

Use civil- or common-law English?

How should you translate terms that don’t exist in common-law English?

When translating a whole piece of legislation, you might be more open to translating non-equivalent terms with a source-language orientation. This entails translating transparently (literally) or using English civil-law terms (if they exist), among other strategies.

Translating with a source focus favours accuracy. But there is a downside. English speakers, even lawyers, might not understand the terms you use. So whatever you do, make sure that the term is clear, describing it (in the flow of the text preferably) if necessary.

A stand-alone piece of legislation can act as a dictionary to itself. So when translating a whole text, you have more leeway to describe terms along the way. Besides, people reading foreign legislation will expect to find unfamiliar terms.

However, a source focus can get you into trouble when translating a snippet from the legislation in another document (e.g. a contract, judgment or letter). Snippets don’t usually provide enough context for readers to get a grip on foreign terms. So in snippets you might want to steer clear of using transparent and civil-law terms as translations if possible.

Take, for example, the Spanish legal term cosa when referring to property, goods or assets (cosa means other things, too).

“Thing”, the literal translation, is not a legal term in English and sounds out of place in legal texts. So when I translate cosa in a snippet — and just about anywhere else — I avoid “thing”. Depending on the context, terms like “property”, “asset” or “goods” usually work.

However, if I were translating the Spanish Civil Code, I might use “thing”. Mainly because if I used different functional equivalents each time, it could feel like I was obscuring a key term that the reader might want to see. As unusual as the term might seem, the reader would understand and get used to it in this context.

Thus, overall, be as target-language friendly as possible and signpost and make clear any foreign-sounding or English civil-law terms you have to use.

Use “shall” for obligations?

You find plenty of language of obligation in legislation. What should you use to express it in English? “Shall”, “must” or something else?

While “shall” traditionally peppers legal writing, current drafting guides advise against using it. They suggest using “must” instead (Australian Office of Parliamentary Counsel, 2013; UK Office of Parliamentary Counsel, 2020).

The traditional style uses “shall” for the imperative. However, the word is ambiguous, as it can also be used to make a statement about the future. Moreover, in common usage it’s not understood as imposing an obligation. Say “must” or “must not” when imposing an obligation, not “shall” or “shall not”. (Australian Office of Parliamentary Counsel, 2013, p. 20)

As I talk about in this post, in legislation you sometimes need to express that a person is required to promise to do something. In these cases, “must” doesn’t always cut it. Use “undertakes to” instead.

For a “gentler form” of obligation, the Australian Office of Parliamentary Counsel (2013) guidelines suggest using “are to” or “is to”. Not everyone will agree with the idea of a gentler form of obligation even existing (you either do or don’t have to do something). But “is/are to” remains a good alternative for obligation in passive sentences in which “must” sounds forced.

Use “shall” for declarations?

The Australian guidelines also advise against using “shall” in “declaratory provisions” (2013, p. 20). E.g.

Don’t write:

    • This Act shall cease to have effect…
    • The meeting shall be constituted by…


    • This Act ceases to have effect…
    • The meeting is constituted by…

Thus, don’t confuse obligations with declarations and use the present tense for declarations.

Use “may not” for prohibitions?

If you’re not vigilant, you can find yourself using “may not” in legislation to express what the law prohibits (aka, language of prohibition). What’s wrong with using “may not”? After all, it rolls off the tongue in these contexts.

The problem is that “may not” can be a source of ambiguity because “it may mean either ‘is disallowed from’ or ‘may or might not'” (Garner, 1987, p. 553), for example:

    • Applications received after September 30 may not be considered by this office.

Is this expressing prohibition or probability?

Confusion like this may be the exception, but it’s good practice to avoid using “may not” for prohibition. Instead, use “must not”. You can also use “are/is not to be” on occasions, particularly when the agent of the action isn’t mentioned.

You may also mean “cannot” when you write “may not” on some occasions. This will be when the meaning is “has no power to” or “is not entitled to” (Australian Office of Parliamentary Counsel, 2013, p. 40).

I crossed out the above paragraph because — despite what the source mentioned recommends — I think we should avoid using “cannot” when translating legislation, even when the original uses an equivalent. In most cases, you will find the purpose of “cannot”, “is not entitled to”, etc.  is to prohibit something, for which “must not” works best.

Use “they” as a singular pronoun?

It is government policy that primary legislation should be drafted in a gender-neutral way, so far as it is practicable to do so. (UK Office of Parliamentary Counsel, 2020, p. 7)

Gender-neutral language should be the default. This is straightforward. But doubts arise with singular pronouns. Should you use “he/she” or “they”? “They” is less clumsy, but it is considered incorrect by some.

A good approach is to use “they” when you have to but avoid it as much as possible by:

    • repeating the noun (not awkward and even helpful in legal texts);
    • changing the pronoun; or
    • rephrasing to avoid the need for a pronoun (UK Office of Parliamentary Counsel, 2020).

UK English seems more tolerant of the singular “they” than US English.

Interestingly, the Australian Government (2013) guide doesn’t mention the “they” option. Although it does suggest recasting to avoid having to use “he/she” and gives using “it” as an option.

Mix “it” and “they”?

You can use “it” in legal texts to refer to a party or an entity. Easy. But you can come to doubt this convention when translating legislation that could apply to both legal entities (juristic persons) and individuals (natural persons). For instance, in legislation that refers to parties to contracts.

In other types of legal documents, you can usually work out what pronoun you should use for a party. But this is trickier in legislation, which, owing to its nature, can be generic and vague.

As a rule, you’d expect a general reference to “the party” to apply to any kind of legal person (entities and individuals), but it’s hard to use “it” if the party is being called a “person”.

In one sentence you might want to use “they” as the (gender-neutral) pronoun for a person. But in the next “it” will seem more appropriate as the party is referred to more broadly as “a party”.

One solution is to use whatever matches the last reference to an entity (usually earlier in the same sentence or just before that). This means swapping from “it” to “they”, but at least the labelling is consistent, and you won’t be using “it” after you’ve just said “person”.

You can also recast so you don’t have to use pronouns in certain places (e.g. using “the party” instead) or make the references plural so “they” works for both. This kind of recasting is easier when translating snippets than whole pieces of legislation.

Replacing the pronoun with a noun is typical of legal style. You can overdo it, though. Australian Office of Parliamentary Counsel (2013) even advises not to follow this tradition and use pronouns instead.

Of course, Australian drafters do have the support of legislation requiring that singular and plural references be interpreted to include each other and that references to one gender be read to include any other (Acts Interpretation Act 1901, section 23).

As translators, we can’t count on any such help short of including a note in the translation. And a note is an option, albeit an intrusive one. It’s easier to manipulate or replace the pronouns.

Recast and use “you”?

You will sometimes be able to use “you” instead of “it”, “they” or he/she” (e.g. when translating company policies).

Using “you” requires recasting. But the document will read better. Plus, “you” is what people expect in English, and using it sidesteps the “they” and “it” versus “he/she” dilemma.

Follow the capitalisation of the source text?

It can be tempting to follow the capitalisation of the source text. However, you should follow target-language convention, taken either from a style guide or what you see in English legislation.

Why? Because it’s what readers expect to see.

For instance, I usually translate “Artículo 128” as “article 128” as this is what you see in English legislation (e.g. “see section 28”).

Follow the section numbering of the source text?

Numbering conventions also differ from language to language.

Spanish, for example, uses written out, ordinal numbers in section headings (Tercero). In contrast, English legislation uses cardinal numbers in figures. Thus, you might want to change ordinals to cardinals in translation.

A problem can arise if the original uses numbers alone without naming the type of section (e.g. “7” instead of “article 7” and “second” instead of “second section”). If you put all the numbering in cardinals, readers might not be able to tell which numbers are for what.

A way around this is to write out section numbers and use figures for articles (i.e. “(section) two” versus “(article) 7”).

UK Office of Parliamentary Counsel (2020) also recommends using Arabic rather than Roman numerals. So this is another thing you might change.

As with capitalisation, the numbering convention is arbitrary. But you have to decide one way or the other. By default, I do what English normally does unless this is awkward.

Translate reference elements literally?

Should you translate internal and external references to other provisions literally or functionally (i.e. using equivalent target terms)?

I go as literal as possible (within reason) to make it easy for readers who want to find the provision referred to in the source documents.

A typical dilemma of this type when translating from Spanish to English is whether to translate ley and artículo as “law” and “article” (literal) or “act” and “section” (English equivalents). There is no right answer, although I prefer the literal option for the reasons I set out in this post.

Get rid of source redundancies?

Both the Australian and UK guides advise getting rid of “the provisions of” in sentences like “under the provisions or this section” when redundant. The UK guide also advises getting rid of “of this Act”.

Should you lose this type of thing in translation when you find them in the source text? If no meaning is lost, I’d say yes. You can justify it as part of translating into a language that provides less signposts.

Make references specific?

You can make a translation more precise by changing things like “in the preceding point” to “in article 129.1” and “see paragraph c) of article 130” to “article 130.c)”. This is akin to adding in nouns when the source text says “in the former/latter” to make sure the reference is clear.

But should you do it? If you want to make your translation as clear and understandable as possible, yes.

Literally translate headings?

Heading conventions differ from one language to the next.

For instance, Spanish legislation often begins headings with De la …., which you’d literally translate as “On the”. But you don’t see this in headings in English legislation. So I’d usually get rid of it. I’d also shorten the heading as much as possible to mimic English style.

Take, for example, the section heading De la rescisión de los contratos in the Spanish Civil Code. You have a few options:

  1. On the rescission of contracts
  2. Contract rescission
  3. Rescission of contracts
  4. Rescission

If this heading were in a document only about contracts, number four might be your first choice. But the Civil Code covers more than just contracts, so number four wouldn’t do.

I’d also discard number one for being unusual in legislation or at least wordy in English.

This leaves two and three. Either would be okay. But it turns out that some of the longer headings of surrounding sections would work better with the “of” structure.

The Australian guide (2013) also advises to avoid noun strings, which is what we would get if we applied the structure in two to the longer headings.

So I’d go with number three to keep the structure parallel for all headings.

So there you have it, some typical dilemmas that crop up when translating legislation. Can you think of any I missed? Let me know in the comments or send me an email.


Australian Government. Office of Parliamentary Counsel. (2013) Plain English Manual. Available in PDF here

Garner, B.A. (1987)  A Dictionary of Modern Legal Usage. 2nd edn. New York: Oxford University Press

Pauino, E., (2013) Legal Certainty and Multilingual EU Law. London: Ashgate publishing

Trosborg, A. (Ed.) (1997) Text Typology and Translation. Amsterdam: John Benjamins Publishing

UK Government. Office of the Parliamentary Counsel. (2020) Drafting Guidance. Available in PDF here

Written by Rob

Rob Lunn is a freelance legal translator based in Spain. He translates from Spanish and Catalan into English.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.