Welcome to the US English Language Guidelines at Unbabel.
If you’re here for the first time, thanks for joining us and taking the first step on your journey to be a superstar editor.
If you’ve been here before, welcome back—you can use the Table of Contents below to find what you need.
Our best editors come from all over the world and from different backgrounds, but one thing they all have in common is they develop, learn, and grow with us. A big part of this is learning what we expect and subsequently checking these guidelines whenever they’re not sure about something.
Read these guidelines carefully, and if you have any queries you can contact us, and we’ll do our best to sort everything out for you.
1. Post-edition at Unbabel
If you’re a new recruit, you might be wondering why we call members of our community “editors” and not “translators”. Well, it’s because there’s a name for the jobs that our editors perform—it’s called post-edition because you’re editing a machine translated text to check for errors and make it sound native—something that machines have always struggled with.
Chances are that if you speak at least one other language you’ll have tried to put something through a translation machine to see how it turns out. Let’s be honest: it’s not always great.
At Unbabel, we recognize that the secret to the best machine translation is having a human check it at the end and make sure it’s perfect, and that’s where you come in! You’ll receive tasks that have gone through some of the best translation machines in the world, and all you have to do is make corrections so that translations sound native and retain their original meaning.
What should I be aware of when editing?
We have a number of systems in place to make it easy for editors to work as efficiently as possible. The three main things you should be aware of are:
Smartcheck. Just like many word processors, we have a system in place to detect errors and suggest corrections—we call this Smartcheck. Smartcheck will let you know if it spots an error by displaying a red exclamation mark on the bottom panel of the editing interface. You can then use your expertise to decide whether to edit or ignore the error detected.
Glossary entries. We use these specific expected translations (displayed in purple boxes) to show you a client’s predefined glossary terms. Unless absolutely necessary you should not alter these. If there is an error, you should report it to us.
Client instructions. Before you start any task, you should always look at the clients’ instructions. These are on the left-hand side of the editing interface, and you must stick to them completely when editing the task. Client instructions always take precedence over the content of these guidelines.
It’s also worth noting that the whole process is designed to anonymize client data so that all parties involved in the translation process are guaranteed security. What this means is that you will find certain information anonymized in the text—one example would be a link that is displayed as URL-0, but there are others.
A word on translation vs. post-edition
Sometimes translators feel like they have to fight against machine translation because it’s too far away from what they’d normally write if they were translating from scratch. As a general rule, you shouldn’t do this for several reasons.
First, take some time to look at the text. Maybe it’s not how you’d have done it, but can you edit it to sound native? It’s possible you’ll have to start over, but this is a rare occurrence with most language pairs.
Second, remember that you could be evaluated on this task, and it’s probable that the more you edit, the more you’re likely to stray from the source. Of course, it’s up to you to judge the text and to what degree it will need editing—but use common sense and remember that your work is not to translate from scratch but to edit the machine output. If you work with what you’ve got already, you’re less likely to produce a target text that doesn’t match the source.
Third, remember that your hourly rate is based on your efficiency—this means how well you can deliver tasks quickly without compromising on quality. Deleting everything and translating from scratch takes a long time, and this could impact your hourly rate.
To sum up, embrace the machine translation where possible and don’t struggle against it. It’ll take a bit of getting used to, but trust us—it’s a great skill to have, and it’ll mean you can work with us much more efficiently as time goes by.
This section looks at core concepts and key issues when writing in English and is designed to help you avoid mistakes and stick to standard conventions when editing.
This section covers key points of English grammar, including (but not limited to): agreement, determiners (e.g., the, a/an), prepositions, tenses, pronouns, and verbs.
A determiner is the grammatical term for something that joins with a noun and determines properties belonging to that noun. The most common determiners in English are the articles the and a/an, but there are plenty of others, for example demonstratives (things that demonstrate what you’re talking about, e.g. this/that) and possessives (my, theirs, etc.).
There are two forms of articles in English—definite (the) and indefinite (a/an). Broadly speaking, definite articles are where you are referring to one specific thing, whereas indefinite articles refer to non-specific items:
Definite: Could you pass me the glass of wine in front of you?
Indefinite: Could you get me a glass from the cupboard?
If you’re translating from a language that doesn’t use articles in this way, for example Polish, you should take extra care to ensure that all of your translations have articles where required, as the machine translation may miss this.
Source text (Czech): Pes je v parku.
✘ Dog is in park.
✓ The dog is in the park.
If the noun is plural and indefinite (i.e. referring to something in general), no determiner is used. This use of a general plural without a determiner is a common English construction.
✓ Polar bears are becoming endangered due to global warming.
One of the most common errors in English translation in this area is the insertion of a definite determiner (the) before a plural noun which has not been previously mentioned. In the example above (Polar bears are becoming endangered due to global warming) the fact there is no determiner there indicates that the polar bears had not yet been mentioned or defined in the text and it would therefore be incorrect, in this context, to write:
✘ The polar bears are becoming endangered due to global warming.
With languages that do have indefinite plural determiners, this can prove to be a problem when translating into English.
Source text (French): Jean est professeur et ses sœurs sont des médecins.
✘ Jean is teacher and his sisters are the doctors.
✓ Jean is a teacher and his sisters are doctors.
Prepositions in English can be tricky if you’re translating from another language into English. It's important to check them and be aware of difficulties that can arise in your specific language pair. Some of the most common examples we see at Unbabel are listed below, but it’s impossible to provide an exhaustive list.
Prepositions are generally grouped into three categories: time, place/position, and miscellaneous.
2.2.1 Time prepositions
Examples of prepositions relating to time (although many can be used for position, too):
On, in, at, since, for, ago, before, by
For prepositions that relate to time, in addition to checking that the corresponding preposition carries the intended meaning from the source to the target text, you should make sure that the tense used in English is correct and not a direct translation of the one used in the source language. This is a common mistake for even accomplished non-native speakers of English.
Source text (German): Ich wohne seit 1990 in Berlin.
✘ I live in Berlin since 1990.
✓ I have lived in Berlin since 1990.
Source text (French): J’habite à Paris depuis 1990.
✘ I live in Paris since 1990.
✓ I have lived in Paris since 1990.
However, remember that even if in a different context it would not change in the source language, the context for the English preposition may change and so too must the preposition in the target text.
Source text (German): Ich wohne seit 17 Jahren in Köln.
✘ I live in Cologne since 17 years.
✓ I have lived in Cologne for 17 years.
Source text (French): J’habite à Lyon depuis 12 ans.
✘ I live in Lyon since 12 years.
✓ I have lived in Lyon for 12 years.
2.2.2 Prepositions relating to position or place
Prepositions usually precede the object which they’re referring to (to the station; with the dog; at the restaurant).
The most notable exception to this rule is when the preposition is found in a relative clause.
✓ This is the book that Paul was reading from.
Formal English teaching in the past insisted that a clause should not end with a preposition, so the construction that was considered correct is This is the book from which Paul was reading.
This construction is however incredibly formal, and for the purposes of translating at Unbabel is not considered necessary.
2.2.3 Adjectival and adverbial prepositions
Prepositions in English can also take the place of adjectives or adverbs in certain situations.
2.2.4 Adjectival prepositions
In cases where prepositions are used adjectivally, they must be next to the noun they are describing.
✘ The receptionist welcomed us into the building with red hair.
✓ The receptionist with red hair welcomed us into the building.
2.2.5 Adverbial prepositions
In some cases prepositions can be used in lieu of an adverb to form part of a set phrase.
✘ We arrived at good time.
✓ We arrived in good time.
There is no rule for this type of preposition, so if you’re not sure, take a second to search for it on the internet. You’ll usually be able to see very quickly which preposition you should use.
2.2.6 Clashing prepositions
English has a lot of phrasal verbs where a preposition forms part of the action. This means that inevitably two identical prepositions sometimes appear next to each other in sentences. This is not grammatically wrong in every case, but it sounds odd most of the time and if you see this it should be avoided, either by changing one of the prepositions (if possible) or rearranging the sentence.
✘ Despite being on a diet, he gave in in the shop and bought some ice cream.
✓ Despite being on a diet, he gave in at the shop and bought some ice cream.
✓ Despite being on a diet, he gave in and bought some ice cream in the shop.
3.1 Contracted verb forms
Contracted verb forms are incredibly common in English. With these contractions, it’s important to strike a balance—no use at all can seem a little uptight and often will sound as if the text wasn’t written by a native, but overuse can seem too familiar in many written contexts.
The use of these contractions corresponds roughly to the register of the text; the more formal the text, the less contractions should appear.
These fall into three main categories:
- Pronoun + to be or to have: I’m, he’s, I’ve, they’ve...
These are acceptable for any text with an informal register, and some limited use in formal texts. This should mostly be limited to use of I’m in first-person texts and customer service emails— anything else in this group should be used with extreme caution.
- Auxiliary or modal verbs + negative: don’t, haven’t, aren’t, can’t...
These are acceptable in texts with an informal register but should not be overused. They must not be used in any text with a formal register.
- Modal verbs in conditional sentences + to have: should’ve, could’ve, would’ve...
The only acceptable use of these contractions would be in extremely informal text, for example a blog article or marketing language. In emails they should not be used.
Double contractions (e.g. shouldn’t’ve) are a feature of colloquial speech and must not be used in any text at Unbabel.
3.2 Picking the right tense
All languages, even those closely related to one another, use tenses to a varying degree, and English uses a peculiar mix of tenses it has inherited from both the Germanic and Romance language groups. It’s therefore probable that some tenses in the source language will not correspond with their English equivalent, assuming that an equivalent exists in the first place.
For example, the future subjunctive, although common in Portuguese, is uncommon in English and translating it directly into its English counterpart would result in a bad translation. Often the answer is to use a relatively simple English tense:
Source text (European Portuguese): Se receber a sua encomenda, por favor envie-nos uma mensagem.
✘ If you were to receive your order, please send a message.
✓ If you receive your order, please send us a message.
It is impossible to produce a list of all these issues, which is why it’s important to know the difficulties specific to your language(s) when translating into English.
4.1 Collective or mass nouns
Collective or mass nouns are almost always plural.
✘ The majority of our business are conducted in Asia.
✓ The majority of our business is conducted in Asia.
However, in the case where the mass noun refers to something plural, in this case ‘customers’, the verb should agree with the plural noun.
✘ The majority of our customers pays with credit card.
✓ The majority of our customers pay with credit card.
4.2 Incorrect translation of grammatical number
There are also a number of common cases we see across all languages where the noun is singular or plural in the source text, but in English it is the opposite.
Source text (Spanish): Las noticias sobre el proyecto están disponibles en nuestro sitio web.
✘ News about the project are available on our website.
✓ News about the project is available on our website.
It would not be feasible to provide a list of all such nouns, but pay particular attention to news and information, both of which are singular in English and have no plural form. Please also be aware that the machine translation engine can get these wrong, so it’s important to check texts before submitting.
4.3 Agreement differences in British/US English
Please be aware that there are some differences between British and US English, namely with company names or similar grammatical forms.
English (GB): Apple are releasing the new iPhone on Monday.
English (US): Apple is releasing the new iPhone on Monday.
As with all variety issues, provided you stick to either standard within the same text there will not be an issue (unless the client instructions give a preference).
4.4 Gender in agreements
In many languages the gender of the person in the sentence may be obvious, in which case you should translate that gender into English, like so:
Source text (German): Der Präsident war gestern mit seinen Kindern in Berlin.
✓ The president was in Berlin with his children yesterday.
If you find yourself in a situation where the gender is unclear, historically you would have used his as a default. This is no longer acceptable, and you should use the gender neutral form of they or its related forms.
✘ A customer called about his package yesterday.
✓ A customer called about their package yesterday.
English is a Germanic language and has the remnants of a complex case system, but in practice there are only a few traces left in everyday language which are normally made clear to students of English very early on, for example the use of personal pronouns after prepositions:
✘ Tom brought his dog with he to the beach.
✓ Tom brought his dog with him to the beach.
The most common error found in these constructions is when two pronouns, or a noun and pronoun, form the subject and one or more object pronouns are used in error:
✘ Janet and me went to the park.
✓ Janet and I went to the park.
Some native speakers often get this wrong. The way to remember how to correctly apply this is to take the other person out of the sentence and see if it sounds natural:
Janet and me went to the park.
Janet and I went to the park.
Here, you can see that the correct form is clearly the second, as opposed to the first; Me went to the park is incorrect English.
In practice, the only other issue we frequently encounter at Unbabel is the use of genitive in English. In singular form, this is expressed using an apostrophe and s after the possessing person or object.
✓ The customer’s details were incorrect.
This can prove difficult in languages that express possession differently, for example using a construction that would translate to of the:
Source text (French): L'adresse du client n'est pas disponible.
✘ The address of the customer is unavailable.
✓ The customer’s address is unavailable.
Although the first translation is not grammatically incorrect, it is, stylistically speaking, a poor choice, and the second is much preferred.
Remember, too, that in the case of plurals the position of the apostrophe must switch, and there is no extra s.
✘ We have three dogs and a cat. The dog’s beds are downstairs and the cat sleeps upstairs.
✓ We have three dogs and a cat. The dogs’ beds are downstairs and the cat sleeps upstairs.
These are abbreviations based on the initial letters of a term (e.g., AIDS, NASA, USA, UK). They are written in full capitals without any spaces, and no periods are used.
✘ Following the agreement between the U.S.A. and the U.K., the U.N. adopted a new strategy to tackle the aids epidemic.
✓ Following the agreement between the USA and the UK, the UN adopted a new strategy to tackle the AIDS epidemic.
5.1.2 Abbreviations of simple words
Abbreviations which are entirely lower case or end with a lowercase letter are generally followed by a period: Mr., Mrs., etc., vs., incl.
✘ Mr Kido, thank you for your inquiry.
✓ Mr. Kido, thank you for your inquiry.
✘ Do you want me to tell you which item I kept vs the ones I returned?
✓ Do you want me to tell you which item I kept vs. the ones I returned?
However, there are some instances where this is not the case, namely with units or measurements.
✘ She cycled 50 km. yesterday.
✓ She cycled 50 km yesterday.
5.1.3 Abbreviations of multiword expressions
Each abbreviated word inside the abbreviation of a multiword expression or phrase (as in Latin abbreviations) is usually written in lower case and followed by a period: e.g., i.e., a.m., p.m.
✘ The recommendation is to log in with some linked account (eg. Facebook) and change the email.
✓ The recommendation is to log in with some linked account (e.g. Facebook) and change the email.
Where a sentence ends in an abbreviation, never add an extra period to mark the end of the sentence.
✘ You must check out before 11 a.m..
✓ You must check out before 11 a.m.
There are two separate uses of the apostrophe in English: possessive apostrophes and apostrophes with contracted verb forms.
5.2.1 Possessive apostrophes
The most common error here is confusion between singular and plural possessive nouns.
The singular genitive (one boy):
✘ The stones hit the boys' head.
✓ The stones hit the boy's head.
The plural genitive (a group of boys):
✘ The stones hit the boy's heads.
✓ The stones hit the boys' heads.
Most nouns are made plural by adding an s or es to the end of the word. However, not all nouns follow this rule (e.g., women, children). Where English has an irregular plural, the apostrophe comes before the s.
✘ The childrens' progress was remarkable.
✓ The children's progress was remarkable.
Compare with the plural noun students, which is not irregular:
The students' progress was remarkable.
5.2.2 Apostrophes with contracted verb forms
In informal writing and everyday speech, it is normal to contract the verb to be as well as most auxiliary verbs (e.g., can, should, have, do, etc.).
She does not have time to finish the project.
She doesn’t have time to finish the project.
I cannot wait to see you.
I can’t wait to see you.
Note that the apostrophe here is written in place of the vowel o in not:
These cups are not intended for hot drinks.
✘These cups ar'ent intended for hot drinks.
✓These cups aren't intended for hot drinks.
Pronouns with have and be can be contracted in informal writing:
It is = It's
I am = I’m
We are = We're
They are = They're
✘ Im happy to help.
✓ I’m happy to help.
Note that we usually avoid using contracted verb forms in formal writing. However, some clients allow it, depending on the audience and context. Revisit section G3.1 Contracted verb forms for further help.
5.2.3 The incorrect, intrusive apostrophe
Contracted verb forms can cause a common error: it's is the contracted verb form for it is and NOT the genitive pronoun.
✘ The government considered it's position very carefully.
✓ The government considered its position very carefully.
✘ The company revised it’s guidelines.
✓ The company revised its guidelines.
When in doubt, read the sentence without the contracted verb form. For example, The government considered it is position very carefully. This sentence does not make sense grammatically, so the apostrophe would not be used.
5.3.1 Places and languages
All place names, country names, region names, and their related adjectives are capitalized. This extends to language names as well (e.g., France, the Arctic Circle, the Italian people, the German language, Chinese food, Tuscan wine, Balinese dances, Greek cooking).
✘ The waiters speak fluent english.
✓ The waiters speak fluent English.
5.3.2 Cardinal directions
Points of the compass are lowercase (unless they are part of a place name, for example, South Sudan).
✘ In the North of this region, there are extensive vineyards.
✓ In the north of this region, there are extensive vineyards.
5.3.3 Months, days of the week and holiday names
Names of the days of the week and names of months are capitalized.
✘ on monday, 5 september
✓ on Monday, 5 September
Months may be abbreviated in official correspondence, table an forms, financial documents or any other chart, calendar or schedule. When abbreviating, use Jan., Feb., Mar., Apr., May., June., Jul., Aug., Sep., Oct., Nov., Dec.
5.3.4 Titles, brands and product names
Use title case in titles, brands and product names. That means the first letter of each word making up the title, brand or product name should be capitalized:
✘ Microsoft office
✓ Microsoft Office
✘ Open file (menu element)
✓ Open File
5.3.5 Institutions and administrative bodies
Names of institutions, organizations and administrative bodies have initial capitals, where other languages use lowercase.
Source text (Italian): Il ministero di turismo
✘ The ministry of tourism
✓ The Ministry of Tourism
Accents (special diacritics) are not used in English, but they may be present in foreign loanwords such as French (née), German (über), among others. In such cases, consult an English dictionary to confirm the correct usage.
5.5 Foreign words
Foreign words and phrases commonly used in English are not translated. If they have entered into English, appearing in English dictionaries, they are written normally: in vitro fertilization; bistro; ratatouille; doppelganger.
Source text (Italian): Suppongo non voglia che venga interrotto il suo tête-à-tête con tuo fratello.
✘ I suppose she doesn't want me to interrupt her head to head with your brother.
✓ I suppose she doesn't want me to interrupt her tête-à-tête with your brother.
Always respect the source text: if numerals are written as digits in the source text, this should be maintained in the translation; if they are written as words, they should be translated to the target language. The only exception is with date formats.
5.6.1 Numbers as numerals
Larger numbers have a comma and no space.
✘ 23 700
Decimals use a period.
5.6.1 Numbers as words
As a general rule, unless specified by the client, write out numbers less than 10 using words; numbers 10 or larger should be written using digits.
If the client asks for numbers to be written as words, compound numbers less than 100 are hyphenated.
✘ There are twenty three restaurants in this area.
✓ There are twenty-three restaurants in this area.
Percentages symbols are written without a preceding space.
✘ 85 %
The slash symbol must be written without preceding or following spaces when used to introduce different alternatives.
✘ US / British English
✓ US/British English
6.3 Currency symbols
Currency can be found under the section for Localization challenges.
A quantity in degrees is written as a numeral, followed by a degree symbol:
7.1: Punctuation marks
Note that, in English, punctuation marks are never preceded by a space.
✘ The story had three parts : beginning, middle, and end.
✓ The story had three parts: beginning, middle, and end.
In a list of three or more items, items are separated by commas rather than conjunctions. The last item in the list is preceded by the conjunction and.
Depending on the situation, you might use an Oxford comma, which is where the conjunction ‘and’ should be preceded by a comma. This can help to clarify the meaning of the sentence and can be extremely useful if used correctly.
✘ I love my parents, Lady Gaga and Albert Einstein.
✓ I love my parents, Lady Gaga, and Albert Einstein.
✘ My usual breakfast is coffee, bacon and eggs and toast.
✓ My usual breakfast is coffee, bacon and eggs, and toast.
In an ordered list where there is no ambiguity, you can leave out the Oxford comma.
✘ For this recipe you will need tomatoes peppers garlic.
✓ For this recipe you will need tomatoes, peppers and garlic. (Without Oxford comma.)
✓ For this recipe you will need tomatoes, peppers, and garlic. (With Oxford comma.)
The colon is used to introduce an element or series of elements that illustrates what has preceded the colon.
✘ In Amsterdam there are a number of museums worth visiting the Rijksmuseum, the Van Gogh Museum, and the Anne Frank House.
✓ In Amsterdam there are a number of museums worth visiting: the Rijksmuseum, the Van Gogh Museum, and the Anne Frank House.
The semicolon is used between two independent clauses which are not joined by a conjunction. However, unless you’re absolutely certain that semicolons are being used correctly in the text, it’s usually better to leave them out or find an alternative.
✓ Call me next week; I will let you know how things went.
✓ Bill is going bald; he is losing his hair.
A semicolon is used in place of a comma to separate phrases or items in a list or series when the phrases or items themselves contain commas or are especially long.
✓ Harry's three favorite bakeries are in Pittsburgh, PA; Phoenix, AZ; and Walla Walla, WA.
The word "however" should be preceded by a semicolon and followed by a comma. When the word "however" is a conjunction, it may be preceded by a comma, but not by a semicolon, and requires no punctuation to follow.
It was clear they could not persuade these people; however, they tried - Here, the meaning is that the subjects tried to convince these people in case, regardless.
It was clear they could not persuade these people, however they tried - Here the meaning is that the subjects tried to convince these people, in whichever way they could
The period (known as a full stop in British English) marks the end of a sentence. It is followed by a single space before the next sentence.
Errors arise when a period is used with brackets. If a full sentence beginning with a capital is enclosed in brackets, the period is within the brackets.
✘ Guests are taken to the hotel by shuttle. (Any guests who miss the shuttle will need to take a taxi).
✓ Guests are taken to the hotel by shuttle. (Any guests who miss the shuttle will need to take a taxi.)
When the words enclosed in the brackets are included within another sentence (even if they form a complete sentence), the period is outside the brackets.
✘ Guests are taken to the hotel by shuttle (the shuttle runs every hour.)
✓ Guests are taken to the hotel by shuttle (the shuttle runs every hour).
7.1.5 Exclamation marks
These are used sparingly in English, to mark emphasis.
An exclamation mark should be placed inside quotation marks or brackets if it is the quote or parenthetical itself that bears the emphasis.
✘ He walked off the stage to cries of "Bravo"!
✓ He walked off the stage to cries of "Bravo!".
✘ He asked if I remembered him (as if I could have forgotten)!
✓ He asked if I remembered him (as if I could have forgotten!).
When you’re translating from a different language into English it is possible that your source text may contain more exclamation marks than you would normally expect. The most important thing to look at is the register of the text, because while you are expected to retain the spirit of the original text, it also may not be possible to keep all the exclamation marks. This is especially relevant if the register is marked as formal.
In these instances you should use your own professional judgment to produce the best-quality text you have.
7.1.6 Question marks
Question marks are used to indicate a direct question.
✓ Is it worth the risk?
They are never used for indirect questions.
✘ He asked himself if it was worth the risk?
✓ He asked himself if it was worth the risk.
A hyphen can be used as a separator between number clusters for clarity, including telephone numbers (0044-01997-042351), ISBNs (978-0-226-10749), etc.
Please note that when working with Unbabel, most number clusters, including telephone numbers, will be anonymized.
Hyphens are also often used when writing compound adjectives.
✓ The council asked for a long-term loan.
✓ The business needs computer-literate people.
Note, since there are few fixed rules about compounds, compound words should be checked in the dictionary.
The en-dash (alt+hyphen) is used in dates. It also can be used in other contexts when meaning to.
✓ Tourism grew in the years 2005–2008
✓ The Rome–Milan train leaves at two o'clock.
✓ The Senate voted 113–67 to repeal the act.
The em-dash (alt+shift+hyphen) is widely used to show parenthetical information instead of commas, colons, brackets, etc. It does not have preceding or following spaces.
✘ It was the most potent image in democracy — the revolutionary idea.
✓ It was the most potent image in democracy—the revolutionary idea.
✓ The match—which had already been interrupted by spectators—was finally abandoned.
✓ The influence of these Venetian painters—Titian and Veronese—was immense.
7.1.10 Quotation marks
Double quotation marks are used in US English. They are used in dialogues to show direct speech and for quoted material.
✓ The teacher began by saying, "Good morning."
✓ What is that poem that begins,"Roses are red"?
Errors usually relate to the punctuation of the closing mark. The question or exclamation marks are placed inside the quotation marks if they are part of the direct speech or the quotation.
✘ I simply asked, "Can you bring me a glass of water"?
✓ I simply asked, "Can you bring me a glass of water?"
They are placed outside if they are part of the surrounding text and NOT the material enclosed by the quotation marks.
✘ What is that poem that begins, "Roses are red?"
✓ What is that poem that begins, "Roses are red"?
Brackets should be preceded by a space and followed by a space if there is no punctuation, but there should be no initial or closing space within the brackets.
✘ The food in this little bistro( which only has five tables )is truly excellent.
✓ The food in this little bistro (which only has five tables) is truly excellent.
Punctuation marks which do not belong to the material inside the brackets follow the final bracket without a space.
✘ When you reach the top of the hill (and if you have enough breath left,) you will be able to explore the ancient fort.
✓ When you reach the top of the hill (and if you have enough breath left), you will be able to explore the ancient fort.
7.2 Punctuation in greetings and closings
English has standard rules for greetings and closings. Whether formal or informal, the initial letter of every significant word in the greeting is capitalized; only the first word of the closure is capitalized. They are both followed by a comma.
✘ dear sir or madam
✓ Dear Sir/Madam,
✘ yours sincerely
✓ Yours sincerely,
✘ all the best,
✓ All the best,
7.3 Punctuation in bulleted and numbered lists
Lists can prove complex in terms of punctuation, and the choice of whether or not to use semicolons and/or periods can complicate matters.
For the sake of simplicity, for Unbabel tasks you should follow the rule below.
If a phrase ending with a colon introduces a bulleted list of elements:
- Don’t end them with periods if all list elements are short phrases (three words or fewer).
- End them with periods if at least one element of the list is a long phrase (four words or more).
8.1 Types of registers
Register deals with the way we use language differently in different circumstances; for example, the level of formality used in the text. It shows how our clients address their customers, and it contributes to the voice of the brand itself. Register may vary depending on the company, the brand, the service offered, the customers, and the target language. Some forms of register include static, consultative, or intimate. We’ll focus more on formal and informal registers, as shown below.
8.1.1 Formal register
The formal register is used in professional, academic, or legal settings where communication is expected to be respectful and restrained. Slang is never used, and contractions are rare.
|Category||Examples||Lexical or Grammatical|
Dear Dr. Dupont,
|Contractions||Not used in formal writing||Grammatical|
|Double negatives||Not used in formal speech or writing||Grammatical|
8.1.2 Informal register
People use this register when they're with friends, close acquaintances, coworkers, and family. Use of slang, contractions, and vernacular grammar is common. At Unbabel, we do not use things like double contractions (e.g., shouldn���t’ve) and double negatives; these are features of colloquial speech and should be avoided.
|Category||Examples||Lexical or Grammatical|
|Contractions||Use of verbal contractions: can't; don't; mustn't; you’re||Grammatical|
Used in colloquial speech, for casual or intimate registers, should not be used at Unbabel.
I can't find my keys nowhere.
I hardly have no money.
9. Localization challenges
9.1 Proper nouns
Proper nouns refer to unique entities including persons, places, organizations, brands, and events. Languages may adopt different rules regarding whether foreign proper nouns should be translated or kept in the original language. When editing a text, you should follow your language’s rules for the types of proper nouns listed below. However, please note that if there is a glossary provided by the client that includes these types of units, you should always apply the glossary items.
With a few exceptions, proper names are not translated into an English form. However, please note that at Unbabel, all personal names will be anonymized when working with some clients.
Source text (French): Bonjour Guillaume, Je serais ravie de pouvoir vous aider.
✘ Hello William, I would be delighted to help you.
✓ Hello Guillaume, I would be delighted to help you.
When names of countries and large cities have an obvious English equivalent, this should be used, but in general, you should leave less globally known place names in their original form.
Source text (Italian): Val d’Aosta
✘ The Valley of the Aosta
✓ Val d’Aosta
With machine translation, errors can easily arise through the translation of place names.
Source text (Italian): Il terremoto dell’Aquila del 2009
✘ The earthquake of the Eagle of 2009
✓ The L’Aquila earthquake of 2009
Global organizations’ names or acronyms should be translated into their English equivalents.
Source text (French): OMS (Organisation Mondiale del la Santé)
✓WHO (World Health Organization).
Government ministries should be translated, even when there is not an exact equivalent in the US or Great Britain:
Source text (French): Ministère de l'Intérieur
✓ Ministry of the Interior
Other organizations should normally be left in the original language, but glossed with an English explanation in brackets.
Source text (Italian): CONSOB (La Commissione Nazionale per la Società e la Borsa)
✓ CONSOB (the National Commission for Companies and the Stock Exchange)
9.1.4 Brands and products
Brand names should always be left untranslated, provided that is the name by which they are internationally recognized (e.g., Sony, Yamaha, Stiga Stjärnsunds).
Source text (Japanese): サムスンの製品とサービス
✘ サムスン products and services
✓ Samsung products and services
Major international events normally have an English equivalent, and these should be used.
Source text (French): Festival de Cannes
✓Cannes Film Festival
Certain international events don’t have an English equivalent and are never translated.
Source text (French): Le Tour de France
✘ The Lap of France
✓The Tour de France
9.2 Acronyms and initials
Acronyms should always be translated when there is an English equivalent. Note that, in English, acronyms are normally written with capital letters.
Source (French): Un financement du SIDA
✘ Sida funding
✓ AIDS funding
9.3 Date format
The machine translation should generate US English-appropriate dates in month-day-year format. However, due to the variety of date formats available, machine translation may not always get it right.
We'd ask you to pay particular attention to the source text before submitting.
Dates may show variation in the following aspects:
- Leading zeros in days: both 09-30-21 and 9-30-21 are fine
- Separators: hyphens and slashes are fine (09-30-21 and 09/30/21 are correct).
- You might come accross dots as date separators (09.30.21), in which case we'd ask you to adopt one of the formats above
- 4-digit or 2-digit year: both 09-30-2021 and 09-30-21 are fine
We strongly advise you pick the combination of the above elements that is most similar to the machine translation visible to you, and that you are consistent throughout the text.
In dates written out in full, use a comma to separate the day from the year.
✘ On October 3 2022 we’re going to Berlin.
✓ On October 3, 2022 we’re going to Berlin.
9.4 Time format
English differs from other languages by using, in general, the 12-hour clock with an a.m./p.m. (ante-meridiem–before midday; post-meridiem–after midday) format. AM/PM and am/pm are also acceptable - for more information see section O1 above.
The numbers are usually written with a colon in US English, and the abbreviations are preceded, as shown, by a space:
5.22 = 5:22 a.m.
17.22 = 5:22 p.m.
8.00 = 8:00 a.m.
20.00 = 8:00 p.m.
Note that the times with :00 can also be written without zeros; in general, try to be consistent with this throughout the text.
✘ Bookings for the same night should be made before 19h.
✓ Bookings for the same night should be made before 7:00 p.m.
✓ Bookings for the same night should be made before 7 p.m.
The abbreviations a.m. and p.m. can also be written as A.M. and P.M., am and pm or AM and PM. The chosen form must be consistent throughout the whole translation.
✘ You can reach us by phone on weekdays from 11 AM to 4 p.m.
✘ You can reach us by phone on weekdays from 11 am to 4 PM.
✓ You can reach us by phone on weekdays from 11 AM to 4 PM.
✓ You can reach us by phone on weekdays from 11 a.m. to 4 p.m.
NOTE: In the 12-hour system, numbers are not always used to express noon or midnight.
✘ The meeting began at 9:45 a.m. and ended at 12:00.
✓ The meeting began at 9:45 a.m. and ended at noon.
Source text (Spanish): No llegamos a casa hasta las 00:00.
✓ We didn’t make it home until midnight.
Measurements should always keep the format of the source text and should never be converted. You should only translate them when they have an equivalent term in your language.
Measurement abbreviations, such as kg, lb, ft, km, m, etc., are generally preceded by a space and not followed by a period (unless they’re at the end of a sentence).
When we write out the word, we use the plural form, while the abbreviation doesn’t have the s.
✘The running track is 2 kms.
✓The running track is 2 kilometres long.
✓The running track is 2 km.
You should always keep currency initials (USD, GBP, RUB, INR, DKK, NOK, etc.), as they are a convention accepted worldwide. The symbol usually occurs before the amount, followed by a space.
✓ USD 2.5 million
Currency symbols ($, €, £...) occur before the amount as well, but they have no following space.
✘ $ 45.54
✘ 45.54 $
✘ € 4
✘ 4 €
9.7 Phone numbers
There is no hard and fast rule for phone number formats. For ease of reading, the country code, dialling code, and area code, if present, should be separated from the rest of the number.
✓ (+44) 121 987 6543
For separating the digits in the phone number, you should use whitespaces as a rule of thumb. Dots or dashes can be used but are not appropriate in all countries, so stick to whitespaces unless a customer has specifically requested it in the client instructions.
✓ 216 732 1578
For US telephone numbers, use parentheses to separate the area code from the seven-digit phone number.
When translating international addresses into English, as a rule, keep them in the original language, with only the country name written in English.
Source text (Portuguese): Rua Passos Manuel, Portugal
✘ Passos Manuel Street, Portugal
✓ Rua Passos Manuel, Portugal
Source text (German): Platz der Deutschen Post, 53113 Bonn, Deutschland
✘ Platz der Deutschen Post, 53113 Bonn, Deutschland
✓ Platz der Deutschen Post, 53113 Bonn, Germany
For US addresses on letter and forms, the standard format is as follows:
457 Cherry Lane #5
Albany, New York 12084
9.9 Idioms and sayings
These can only very rarely be translated literally. Occasionally there will be an equivalent English saying, but otherwise, you need to simply express the intended meaning.
Source text (Italian): Dobbiamo ingoiare il rospo.
✘ We have to swallow the toad.
✓ We have to bite the bullet.
✓ We have to put up with it.
9.10 Literal translations
In addition to the normal problems of literal translation, in French correspondence for example, courtesy phrases are often used, which should never be translated literally into English. Instead, the corresponding English phrase should be used.
Source text (French): Je vous invite à consulter votre courriers indésirables.
✘ I invite you to consult your undesirable couriers.
✘ I invite you to check your spam folder.
✓ Please check your spam folder.
✓ I suggest that you check your spam folder.
9.11 Inappropriate word choice
Machine translation, even when of a high standard, sometimes uses the first dictionary definition of a word, which is not always appropriate in the target language.
(The context here is a parcel being returned to its sender. The parcel will be collected by a courier, who will provide the return label.)
Source text (Italian): al momento del ritiro del pacco
✘ when the package is withdrawn
✓ when the package is collected
10. Useful online resources
These bilingual dictionaries are helpful because they allow editors to see words as they are used in context through external sources and forums. However, they may not be useful for every language. For example, bilingual Spanish speakers may use this resource with ease, while Portuguese speakers have found errors in the entries.
10.2 English verb conjugator
10.3 English grammar guide
10.4 English dictionaries
Merriam Webster is the quintessential dictionary for US English. Although less used, The American Heritage Dictionary of the English Language is also well respected.
A thesaurus is used to find synonyms or antonyms for words which can be more precise and descriptive; it’s useful when deciding the best word choice. However, it’s important to not overuse this resource. If you use a thesaurus to substitute a word, the word must be correct, and you must understand it well enough to know whether it’s correct or not. When in doubt about a word, don’t use it.