Indeed, both translation suppliers, as well as clients of translation services providers have vastly different views on the issue of the desired versus feasible speed of professional translation and localization.

While it's not easy to answer the headline question in a one-liner, let's try to break it down. Basically, the translated content's nature speaks for itself. Usually, the more time it took to write, the more time it takes to translate. Shakespeare plays, important medical research papers and a kitchen timer manual are just about as different as would be their translation processes. There are many types of translation areas that are very specific and thus require specific know-how, translation skills, and experience.

Literary vs. technical translation

So, the deepest divide is the nature of translation, or better said whether we talk about a literary translation that borders with an arts discipline and has artistic value added, or the technical translation fields where the content "just needs to be translated correctly".

Literary translation of a Shakespearean screenplay would require a highly skilled multilingual literate (yes, usually only one to translate the entire piece of work) to localize all nuances of a foreign language to fit the structure, while preserving the original meaning, that is including the artistic value of the content, too. Needless to say that each author has their own style that cannot be just transformed into ones and zeros as in technical translation.

Depending on the volume of content, the literary translation process could take even up to several months, at a speed starting from pages a day to a chapter or two. A careful guess would be approximately 1-2,000 source words/day.

Technical translation of highly specific non-literary content(e.g. medical translation, legal translation, financial translation – texts such as research papers, medical or engineering documentation, etc.). These fields of translation, albeit highly demanding on translators' expertise and a bit similar to art books by their complexity, can employ various strong computer-assisted tools (CAT tools) and well prepared multilingual databases such as translation memories and glossaries to support and eventually speed up the translation process. However, even with good computing power at hand, highly-specific content uses also highly specific terms and expressions. And even though a technical translation of a high-profile content enables (unlike the usual practice in literary translation) a team of translators to collaborate, the translation process can slow down to less than a page per hour. In general, for this kind of translation good translators can produce about 2-3,000 source words/day. General (technical) translation

For the general text (including guides, sales and promo material, website content, etc.), translators supported by CAT tools are usually capable of working at a speed of 2,000 source words/day and more. Such projects can also be split into a team of more linguists, which can then multiply the speed.

Express translation?

Maybe you wonder how long it would take if you would need a professional, high-quality translation of your content urgently? At idioma, we are able to handle smaller projects (up to 200 source words) within 4 working hours (CET), with no minimum fees. If you wish just 5 words translated, we charge you only for those 5 words.

And if you happen to need express translation at the moment, just click here.

There were times, and it is not so long ago, when not even Europe had a unified calendar – not to mention the world. And although the IT revolution made us unify most of the information to 0 and 1, including all everyday thing, calendar dates can still turn into a real pain when it comes to localization.

Calendar dates formatting

There are various formats that different languages and cultures use for writing dates. The reason for such usage of the specific formats are usually historic and cultural, but some are also driven by technical development. The calendar dates can vary as follows:

• Order of date components (e.g. day-month-year = little-endian; month-day-year = middle-endian; year-month-day = big-endian) - the most popular in the majority of countries around the world is the day-month-year format, mainly due to the Western religious and legal customs of writing dates (e.g. the 1st day of November, Anno Domini 2016)
• Usage of leading zeros in days and months (e.g. 01-01-2016 vs. 1-1-2016) – German-speaking and German-influenced regions, for instance, tend to use
• Separators like hyphens, dots, etc. (e.g. 01-01-2016, 01.01.2016, 1 January 2016, 1. January 2016 or 01/01/2016)
• Year format (e.g. 01-01-2016 vs. 01-01-16)
• Numeral type usage – Arabic vs. Roman (e.g. 1. XII. 2016 vs. 1.12.2016)
• Months name usage (months can be written down using both names and numbers, e.g. 1.1.2016 vs. 1.January 2016)
• Other language or cultural specifics (e.g. 1st January 2016 in English, or adding AD (Anno Domini), or CE (common era) to the date)
• Reversed day and month this is a popular format used only in the United States and often a default settings in many computers, e.g. 01-31-2016 for January 31, 2016.

There is also an ISO 8601 standard for data elements and interchange formats, that works with YYYY-MM-DD format.

Time zones matter in dates localization

Not only the formatting, but also timezones need to be taken into consideration, based on the observer's view. This can be rather tricky with important historical dates, where e.g. the attack on Pearl Harbor, generally known to be December 7th, 1941, actually took place on December 8th in Japanese time.

Translators can be faced with many file formats that require translation, from simple text files to complex DTP files. These more "exotic" files, such as .mif (FrameMaker file format) or .idml (InDesign format) sometimes bring hell on earth for translators trying to import texts from such files into their CAT tools.

In particular, one of InDesign's neat features – layers and conditional text – allows users to create different versions of a document. However, the presence of more layers or text conditions in an InDesign file may easily become a confusing factor when importing text into a CAT tool. Text may become "lost" at various stages.

InDesign: Where to find relevant text

Let’s take a look at some InDesign features that might contain relevant (and sometimes hidden) text:

LAYERS

An .idml document can be organized by placing text (and objects) onto different layers. It's possible to show and hide those layers to display relevant text.

TIP: To show and hide the layers palette in InDesign, go to Windows > Layers (or press F7). With the layer palette displayed, layers can be hidden and unhidden by toggling the “eye” icon on and off.

CONDITIONAL TEXT

This is content within a document that is meant to appear in some renditions of the document, but not other renditions.

In InDesign it's possible to create text conditions and then apply them to text just like it would be done with character formatting.

TIP: Similarly to layers, certain text can be shown or hidden by toggling conditions on and off. In InDesign, go to Windows > Type & Table > Conditional Text to show the Conditional Text palette. With the layer palette displayed, conditions are hidden/unhidden by toggling the “eye” icon (just like you’d do for layers).

PASTEBOARD

There is yet another place in InDesign where irrelevant text may have been “parked” by the author of the document – the pasteboard. The pasteboard is that blank area around InDesign layout pages. Make sure to clear the pasteboard, otherwise any text that is found there may end up in translation!

One last useful TIP: Translation companies and translators should ask document authors to hide all unnecessary layers and conditional text in InDesign documents that should be translated. This will allow the translators to focus only on the intended content, which saves clients' time and money.

Everything evolves. 30 years ago the translators worked on typewriters, now we work on computers with special translation applications that offer a lot more than delivery of translated text. Today, the delivery includes almost ready-to-publish text that hardly need any reformatting. This is becuase in modern translation, text from source documents is imported into special translation memory applications (for example, we use our in-house developed tool iQube at idioma, which we offer for free to all the translators and linguists that work with us). And besides importing the source text, a translation memory product also imports text formatting. This formatting includes standard commands such as font and font size changes, variables, cross-references for e.g. indexes and pictures, etc. The formatting commands are referred to as tags, and when placed correctly they decide how the translated text will look like in final format.

In most cases, the translator places the tags in the corresponding places in his translation so they more or less match the tag placement in the original. For standard text, this is a process that does not take much time – in fact, the process is highly automated and in many cases tags are put where they should be without human intervention.

Too many tags = loss of concentration

While convoluted formatting commands get compressed to dense, single tags, they can still cause a lot of frustration and make the translation work extremely tedious. There are cases where text in an original can contain so many tags that the text segments can hardly be translated, while trying to simply understand the text can be a challenge (as seen on the screenshot below).

Extreme tag example: Pink tags represent Bold On and Bold Off. Blue tags are redundant with letter spacing info, which is not an issue in translation.

In some cases, especially with OCR text or text that has been edited a lot (in e.g. Word or originate from RTF files), such tags can’t even be seen with the naked eye. This issue can therefore be easilly missed by clients submitting text for translation. However, the tags show up in the translation environment – and they can create such a mess that translators lose their concentration and make mistakes they would normally not make (just try to read the text on the picture :) ).

Cutting down on tags improves translation

So especially with these kinds of files that allow you to get (sometimes excessively) creative in terms of formatting, it is a good habit to apply a general font, style or neutral text format to the texts before submitting the document for translation. This will remove most unnecessary tags and only keep those that are relevant.

The translator can then focus on his translation, choosing the right words in the context and making sure the message in the original is passed on to the foreign reader.

At idioma, it is a habit of ours to inform clients when they submit documents with tag clutter. We inform how tags can be reduced and we see it as a way to ensure quality is consistently high and the translator can do what s/he is actually tasked with: translate instead of doubling as a layout person.

Dutch is a West Germanic language primarily used in the Benelux region and an official EU language. It is closely related to other West Germanic languages (e.g., English, West Frisian and German) and somewhat more remotely to the North Germanic languages. Because of former ownership of colonies on the African continent, Dutch has considerably influenced Afrikaans, one of the official languages of South Africa and the most widely understood language in Namibia.

Nowadays the language is spoken by 24 million people around the world, primarily in the Netherlands, but also in Belgium, where Dutch is one of the official languages, next to French and German. The Belgian version of Dutch is often referred to as Flemish (Vlaams), which covers a group of Dutch dialects with slight differences from standard Dutch. But what are the differences exactly?

Difference between Dutch and Flemish

Flemish, or Vlaams in Dutch, is the standard Dutch variant spoken in the Belgian region of Flanders, with approximately 6.1 million speakers. It includes several dialects, all of which are interrelated with the southwestern dialects of Dutch. The main differences are in pronunciation and frequency of certain words. Because certain words (around 3-4,000) are more frequent in Belgian Dutch, many refer to the language as Flemish, however the words are indeed part of standard Dutch. These different Flemish-preferred expressions are often considered as "old-fashioned" in Dutch. A slightly more old-fashioned sense of Flemish in comparison with Dutch is underlined also by a more formal tone of communication among speakers of Flemish. While Dutch people tend to switch from formal to informal tone rather quickly, Flemish speakers use more formal expressions (for which they are sometimes considered cold or unpleasant). However, there are no spelling differences between the Dutch language used in Belgium and the Dutch one used in the Netherlands.

Alphabet

In addition to the standard English alphabet, Dutch ends with (…) X Y IJ Z. The alphabet is shared also with the Flemish dialects.

A B C D E F G H I J K L M N O P Q R S T U V W X Y IJ Z

a b c d e f g h i j k l m n o p q r s t u v w x y ij z

Translation Memory products constitute an excellent aid for translators and help them to translate bigger volumes in less time. However, there are dangers with using such products also. Poor memories is one thing, but translator fatigues is another that can lead to serious issues and very unhappy clients, especially when working with fuzzy matches.

Some of the more critical errors can be prevented with a so-called opposite translation check.

What is Opposite Translation?

Opposite translation happens when a word gets translated with the opposite meaning, e.g. if Up gets translated as Down, In as Out or On top with Under. Suppose, for example, that a German translator gets tired and misses changing e.g Rechts (right) to Links (left) in a text segment, the outcome could very well be a serious mistake in the final document. The second language reviewer should detect this error, but human error is something that’s difficult to completely stamp out. To protect against this, translators and language reviewers should be warned whenever a suspicious opposite translation is detected.

At idioma we use our in-house developed translation platform, iQube, with support for numerous file formats and a built-in Quality Assurance checker, to alert translators in such cases.

When the meaning becomes completely the opposite

This iQube semantic feature can detect opposite translation of the most common terms and supports around fifty different languages, in all imaginable combinations, and it includes warnings for wrong translation such as: - Under instead of Over - Before instead of After - Disassemble instead of Assemble – and many, many mores – and of course also in reverse. This feature of our iQube translation platform is unmatched by any other product in the market. One of its outstanding features involves advanced morphology to detect opposite translation of critical terms. The Opposite Translation check complements all the other 50+ Quality Assurance checks to further improve quality of translated text.

Unparalleled Quality Assurance

The Quality Assurance checks are part of a mandatory workflow process in all translation projects with 3-stage QA checking: by the translator, by the language reviewer, and finally by in-house QA staff. The checks have cut down significantly on unnecessary human errors, especially those that involve fuzzy matches taken over from translation memories.

This allows us to focus on providing uncompromising high-quality translation, where a full TEP process (Translation – Editing – Proofreading fully complying with the DIN EN 15038 standard) is mandatory for all the texts we handle – all this at very competitive rates.

To learn more about our services, please contact our project managers.

As translation professionals, we take many things for granted when we write and translate. While many of our clients have developed different writing rules for different languages in so called style guides*, which we respect and actively use, some things are pretty obvious to anyone though. Even children learning how to write learn from the start that sentences begin with a capital letter and end with a full stop. Bigger children learn later on that where there must be a pause in a speech, a dash should be used. And here not even adults are always sure, or even aware of the problem: what kind of dash to use?

Do you know your dash?

In fact, there are three types of dashes:

1) The em dash—used for parenthetical thought—is as wide as a capital M and is used mostly in US English. If you are really picky, the em dash should be surrounded by half or quarter spaces, but many ignore this.

2) The en dash – which is as wide as an N – is often used to pause in a sentence and create emphasis. It is also used in ranges, e.g. “pages 8–10” or “ages 2–5”. Some people also use them to indicate negative numbers, like –15°C.

3) Lastly we have hyphens, which are used to hyphenate long words at the end of a line. They can also be used to connect words, in English to make them easier to read as in “state-of-the-art”, and in e.g. Germanic languages to connect them with loan words or proper nouns (Ford-Partner in German, for example).

Other cool stuff that people in typography are familiar with are soft hyphens. These can be used to “softly” hyphenate words at the end of a line. Try typing a really long word that will not fit at the end of a line, then place a soft hyphen in the middle and see what happens – you can create one by pressing "Ctrl" and "-" at the same time.

Nice, huh? Relying on details like this may seem too extreme, but in terms of localization, this is what (among others) makes a difference in quality and know-how in translation.

...and do you know your quotes?

Not only dashes/hyphens, or styles of writing numbers – also quotes and their usage differ among languages. English uses sixers and niners, like “abc”, in France they use « abc », Germans and Czech write „abc“, while Danes write »abc« and Swedes like niners and niners, like ”abc”. There are numerous other types, and as a translator, it is important to know which one is correct.

For example, in Japan, the written language and customs are drastically different. Apart from being able to write from up to down with “line columns running right to left”, thousand separators do not exist. Instead people separate digit groups in units of ten thousand. They also do not have dots as full stops, but instead use small circles, and quotes become「abc」.

If there are client preferences, then these become exceptions and the translator must be alerted (which our translation platform iQube can do automatically).

Your Solution? Style guides

Writing rules for different languages are explained in great detail in style manuals, such as the Chicago Manual of Style. Such manuals are actually interesting reading with a surprising wealth of information. However, in the end, it is hard for an experienced translator to remember and keep with all the different rules that apply, and to this end we rely on condensed forms of the manuals, commonly referred to as Style Guides. They contain the basic rules on text presentation and are either prepared by individual clients, or, if they are lacking, we are happy to help create them based on general rules and local preferences. For more information, visit www.idioma.com!

*These style guides can be loaded on our iQube translation platform. Each time a text segment is opened and an issue covered by the style is detected, this is pointed out to the translator and subsequently the reviewer.

At idioma we translate into many different languages, over seventy at the last count, and in many different combinations (last we counted the permutations, it was over five thousand). Some of these languages are different variants of the same basic language. Sometimes they are very similar, at other times quite different.

Flavored by authorities

For example, for Norwegian, two flavors exist: Bokmål and Nynorsk. Bokmål is by far the most common variant in Norway and is used in the private and commercial sector, while Nynorsk is used mostly in some western regions and in public administration. The two dialects differ considerably from each other.

Then for Portuguese, there are also dif­ferent tongues, and the language used in Portugal differs from the one used in Brazil, for example. In spite of the fact that a language reform has been signed into law (in Portugal in 2008), in reality there are quite many differences. Schools in Portugal now teach the new, standardized Portuguese language, which means that in a global perspective spelling and grammar should become standardized. Most newspapers and magazines in Portugal have also adopted the reform, which officially must have been applied by latest July 2014. The biggest obstacle, however, seems to be the choice of words. Especially in technical writing, there are numerous cases where Brazilians prefer different terms than those used in Portugal, which is one reason why Portuguese for Portugal and Portuguese for Brazil will most likely continue to coexist for quite some time.

Spanish is another case in point where differences exist. Spanish in Spain tends to be quite modern with development in a different direction from other Spanish tongues. Most of the Spanish dialects in South America are quite conservative, while the dialects used in Mexico and the Caribbean are influenced by their proximity to the United States.

Does it matter in localization?

Many times, it is more important to know the target market than the language itself or whether a document will be used in many different markets. In Belgium, for example, people in the north speak Flemish, a variant of Dutch. This dialect can’t really be called a different language as spelling and grammar are the same as for standard Dutch. Here the difference is rather in the frequency in the words used, although all words exist in standard Dutch as well. However, it can make a surprising difference when it comes to, say, localization of websites.

Similar to Flemish, the German vocabulary used in Austria and Switzerland is also common, however, the preferences, especially in Switzerland, are many times for different words than those that are commonly used in Germany. Additionally, in Switzerland the German character “ß” is not used, and instead people write “ss”.

In our work, we come across the issue of language flavor daily. We of course translate into the various dialects mentioned above, and many more, and we will be happy to help with issues regarding which language or dialect to translate into.

We are often approached to provide guidance on what is the most common way of using decimal dividers and thousand separators for the various languages that we translate into.

While everyone is aware of US English using periods for decimal division and commas to separate thousands in big numbers, the issue is somehow obscure when it comes to all the different lan­guages used in Europe.

Diacritical mayhem

In US English, the value of pi is 3.14 while a million is written as 1,000,000 with comma separators. This system is also used widely throughout Asia and in almost all English speaking countries. In Spain, pi is written like “3,14” and everyone would like to win “1.000.000” Euro in a lottery. Many other European countries apply similar punctuation in numbers, but there are exceptions. In Germanic languages, i.e. German, Dutch, Danish, Norwe­gian and Swedish, the decimal comma is also standard and pi is written as “3,14”. However, while a million can be written like what is the custom in e.g. Spain, there is now a general trend to instead use nonbreaking spaces and people like to win “1 000 000” Euro instead on the lottery. To complicate things, for the Germanic (and also Slavic) languages there is also a general preference to not use any divider in single thousand numbers. So you would e.g. pay “€1500” for a very good bicycle and “€9000” for a decent car. If you add German value added tax of 19%, the final price becomes “€10 710” (yes with a space) for the same car.

Solution? idioma QA style sheet

This trend of omitting the separator is also picking up in many other European countries, and it is a commonplace practice today. We have learned that many of our clients are not aware of this. As a result, we have developed special QA style sheets in which we have recommendations for all the 70+ languages we translate into.

Clients can accept these recommenda­tions or enter their own preference for the various languages that projects should be translated into. It is even possible to enter non-standard practices, e.g. to omit all dividers in thousands and larger, for example when translat­ing very technical documents where numbers should simply remain the same in disregard of what language they are translated into. The information in the style sheets is passed on to our translators and proofreaders as ‘mini rules’ so they can adhere to your pref­erences while handling your projects.

Please contact our project manager for more information on these style sheets and how you best can use them to your advantage.

Translation is not an easy task. Many people say that the best translation is when translated text exceeds the quality and understanding of the original text. We fully agree, but to get there is an uphill battle and a struggle unless we know what we are dealing with. In reality, there are hundreds of different ways that you can combine words to make up an intelligent expression that reflect the original meaning without getting into wordiness or veering too far off subject. We do this as a matter of course daily – actually round the clock.

Translating for industries is a continuous undertaking, and we do it in close to a hundred languages. We use translation memories, reference resources, and then we use glossaries – either ones we have made ourselves or those provided by clients. All of this helps, but there are times when we are at a loss, and then we usually ask.

Context matters

Many of our assignments are additions to existing documents, where we translate out of context. This can be a nightmare to a professional translator.

Once in a while, we come across terms that simply cannot be translated unless we have references. As in this title “jack” is a good one. It has so many meanings, especially when it appears as a lone item. Even my good friend Jack agrees.

Many Jacks 1 Jack Jack

But there are many others. When clients ‘help’ and prepare projects in e.g. Trados, we often come across terms like “No” that can mean both Number and the opposite of Yes. It is of course also a Japanese style of theater, mostly performed by men, and it has even more meanings, e.g Nitric Oxide. Here we try not to guess, and we use the context to try to figure out the meaning, but it is undeniably a challenge.

Don't risk. Get reference.

This is why we always emphasize use of references. Original files as PDFs are most helpful, especially with pictures. Just think of another simple term like "cart". This being a reasonably old term, it has numerous different translations in any imaginable language. It can refer to a shopping cart, the horse driven variant, or something small and ‘fast’ that kids love for downhill races and band-aid use. Grandpas and other fellows use it on the golf course, some use it as a beverage carrier, and in Africa you can go on a Rhino safari in a Cart.

The above is a typical translator dilemma. For the Rhinos, you definitely need to emphasize some protective elements when referencing the “cart”; if you are dealing with a grocery store, you simply use the standard “cart”.

The core of this issue is reference. All translators need good reference that supports the text s/he will translate. Glossaries are a good help, a translation memory too – anything less is a risk of stating something imprecise or downright wrong.

Translating from and to many languages brings about lots of interesting facts, increases curiosity and reveals things most people would hardly think twice about. Well we do, so why not share some the know-how?

For instance, we regularly work with translation into English – who doesn't, to be honest :) – and handle both British English and English for the U.S. market. While it is common knowledge that U.K. and U.S. English differ slightly in spelling and also between certain expressions, the two tongues also have very different approaches to the use of Initial Capitals, i.e. the custom of capitalizing the first letter of every word in headers and sub-headers.

The British keep it subtle, Americans Like it Big

To roughly sum it up, people generally favor the use of initial capitals in the U.S., while in the U.K. there is a tendency to avoid them. If you open a newspaper from the U.S. and one from the U.K., this becomes obvious (well illustrated by the picture on the right). To a translator, the issue is trivial, but if you publish documents for English readers in general and there is no target market, it may be worth to reconsider the use of initial caps in headers.

Avoid international Caps schizophrenia

Having noted that Americans like to use initial caps in titles, while the British try to avoid them, for International English we recommend not to use initial caps, because it makes it difficult to balance the heading levels that should have initial caps and those that should not. It is difficult to keep respecting the rule and even more difficult to unify all headers throughout documents and between projects. Further, when writing International English text for a worldwide audience, it is easy to make mistakes if initial caps are used, which is another good reason for advocating the British preference. As to the general opinion of the British that initial caps usage appears ugly, we won’t comment. But let's admit that caps overuse can confuse and distract the reader, a fact that eventually will affect the level of understanding the context and slow down the reading speed. You really don't want your target text to appear that way to the reader – at least not if we talk promotional material.

When to use initial caps?

We do, however, agree with and encourage the use of initial caps in proper names, including product names and to some extent special part names, as well as in key phrases, e.g. catch phrases. This includes also titles (e.g. on book covers - as on the picture on the right).

The easy way out

If you are really stuck and can't make up your mind, there is always an easy way out: Simply capitalize all your headers and reduce the point size to avoid them dominating the content.

For other translation and localization tips, language facts and curiosities, keep tracking our blog!