e2f blog


We’re hiring!

by Peter Corless

How would you like to work for e2f? We’re hiring!

Translation, Localization, Linguistic Quality Assurance (LQA)

We currently have multiple open positions and immediate needs for part-time and full-time staff in translation, localization, and linguistic QA for the following languages (and target markets). Are you proficient in one (or more) languages?

  • Indonesian speakers (full time, a couple months or more)
  • Canadian English speakers (full time long term)
  • Australian English speakers (full time long term)
  • Finnish speakers (full time long term)


  • Malay (Malaysia)
  • Danish (Denmark)
  • French (Canada, Switzerland, Belgium)
  • Dutch (Netherlands, Belgium)
  • Spanish (Spain)

All other languages and dialects also welcome to apply, as we have ever-changing client needs.

Voice-Over Sample Casting Call

e2f has an in-house full-service multimedia studio. We have clients that need voice-over work done for games, applications, videos and corporate presentations translated into a variety of languages.

Whether you speak English or a foreign language, we’d love to add your voice to our growing portfolio of available-on-demand voice-over artists and actors.

This session will be an uncompensated voice sample recording. If you are selected in the future for use by a client, your work at for that project will be professionally compensated.

We hold our casting calls in our San Jose offices. However, no drop-ins. You must contact us to be scheduled beforehand for a recording session. The session will last approximately 15 minutes. Contact us to know when we’ll hold our next voice auditions!

Game Testers

We also are building out a team of game playtesters. These will be short-time gigs of few hours for 2 weeks.

To Apply

See our current ad on Craigslist, or view our Employment page.



Wearables TechCon in San Jose

by Peter Corless

This week, the Wearables TechCon, plus Internet of Things TechCon were held in San Jose’s McEnery Convention Center. e2f attended to see what was new in these exciting fields.

There were a number of emergent trends at the show:

  • Devices are going from concept to market in incredibly quick time frames.
  • Everyone’s in the game: the field of vendors span the gamut from established players to brand new startups.
  • Companies are already making profits.
  • Standards matter.
  • Quality matters.
  • Usability matters.
  • Design matters.

One of the most well-designed products spotted at the show was the new Nut 3 smart tracker (pictured above). This small device, which you can slip into your wallet, attach to your keychain, or tether to just about any other prized possession uses Bluetooth technology to keep in touch with your phone. If you go past a certain range of your Nut-connected item, your phone will beep, and you can use the Nut phone app to track down your item. Handy to prevent you from losing your wallet or keys!

Beyond its pragmatic purposes, the Nut 3 sports a beautiful new aesthetic form-factor. The design invokes the organic curves of rippling water or a Zen sand garden.

Nut’s Michael Tian used a successful Kickstarter campaign to get the new model to market swiftly. Hundreds of backers participated.

Nut’s phone-based app, available for iOS and Android, has already been translated into seven languages.

Vendors, overall, were receptive to the idea to make their devices and supporting apps globalization-ready. It was generally agreed that device designers would be well-served to ensure localization and translation were not left as last-minute additions in go-to-market planning.

How about at your organization? How far up-front in your strategy, design or conceptualization phases do you put localization and globalization engineering? We’d love to hear! Email us at projects@e2f.com.


No, not “Whatever” — It Pays to Proof

by Peter Corless

Last week we posted about a poor Welsh translation made by the chain B&Q reported on the BBC. Some might argue the incident doesn’t matter to them because Welsh isn’t a really a broadly-used language internationally. Focusing on the language is ignoring the point. Our point is that a local business used Machine Translation (MT), with all the best of intentions to reach a local audience, yet without doing even basic proofing or editing. It is unsurprising it blew up in their faces. Which harms the brand in all languages, including English.

Today, it’s the Daily Mail’s turn to spot poor results of what was likely Machine Translation. In this case, in the world’s most spoken-language, Chinese, and also, its rendering into English. Their mocking article, whose very title isn’t particularly business-appropriate, focuses on restaurateurs in Shanghai.

The commonality with the Welsh story is this: small business owners are not known for having large budgets for translation, hence the attraction to using free and open web tools, like Google Translate. Yet it might pay small business owners to submit machine translations for proofing and editing to avoid public embarrassment and shaming. And, in cases where you have specialty items and idioms, maybe it pays to have a full human translation done. Such a cost can be seen as an investment up-front to save from the potential of lost business and pain of embarrassment.

One example from the Daily Mail’s article has to do with the Chinese term “随便” (suíbiàn) which an unfortunate Shanghai menu had translated into English as “Whatever.” That is indeed a quite common translation. Though it is unlikely you ordered a plate or bowl of “whatever.”

There are multiple renderings of this word, from “casual” or “arbitrary,” to “as one pleases” to “wanton” (that is, “careless behavior;” not to be confused with the Chinese food wonton). Mandarin House has a whole article about the term, which can range idiomatically or contextually from browsing while shopping (“only looking”), to “you choose for me,” to “make yourself at home” and “please help yourself to eat.” So it can range anywhere from the blithe to the pejorative in meaning.

The subscript below the translation reads “等答巢料” (Děng dá cháo liào), the latter half of which (“巢料”) infers some sort of “nesting material,” so it doesn’t seem to be much help.

So what does it mean in this case? This is something a computer would have a hard time to answer, or even a human who had no context or contact with the staff, but a quick chat with the restaurateur or cook would be able to answer these questions in a heartbeat.

While it is easy to mock and deride “Chinglish,” this happens in many linguistic contexts. An actual solution to the problem of poor translations would be more emphasis on available services for translation and localization for small businesses here and abroad.

What are your thoughts? Have you ever run into situation that would have been immediately improved with a quick editorial or proofing job? We’d love to hear your stories. Email us at projects@e2f.com and let us know.

Common Sense Advisory logo

e2f in CSA’s Top 100 Global Language Service Providers List in 2016

by Peter Corless

The market research firm Common Sense Advisory (CSA) produces a comprehensive annual report on the state of the Language Service Provider (LSP) market. This year, for the 2016 edition, e2f has entered the august list of Top 100 Global LSPs!

e2f-color-logo-bitmape2f placed #97 in the global list. For North America, we were even more prominent, placing at #28 of of the Top 40 LSPs.

Some of the key advances we made in the past year include:

Stay tuned. We have more exciting news to announce in the second half of the year!

While we have come a long way over our company’s history, we’re just getting started!. Here’s to every one of the customers and partners of e2f, for it is your successes that allowed us to achieve these rankings. And here’s to 2017 being an even more amazing year!

Flag of Wales

Rydym Wedi Symud — “We’ve Moved!” (but not that far…)

by Peter Corless

Sometimes you can tell an automatic translation job when words are obviously mistranslated. A short word, for instance, can be mistaken for an acronym or abbreviation. This is especially true for the common English word “us,” which can either be a pronoun (the objective version of “we”) or, especially if capitalized, can be an abbreviation for “United States.” The translation should be contextually obvious to a human. To a computer? You might need to train your model more.

This was apparently the case when a Welsh store, a B&Q at the Parc-y-Llyn Retail Park in the town of Llanbadarn Fawr, Ceredigion, posted a sign about closing the location and redirecting their customers to the closest alternate store, as reported by the BBC. The sign seemed to be redirecting people inappropriately across the Atlantic!

In English, it was likely originally “Find us at…” because what follows is an alternate address and website of the company. It might have been rendered as “Dod o hyd i ni ar…” But instead, the sign read, “Dod o hyd at Unol Daleithiau…” When translated back to English, that means, “Find at United States…”

A few problems to note. Obviously “us” (the pronoun) was mistranslated as the Welsh version of “United States.” Yet also note the preposition “at,” which was not translated at all, had a strange transposition in front of the direct object “us.”

The sign will be replaced, but the embarrassment is now historical, due to the BBC’s coverage. So consider this example for your own future. When using machine translation, it may actually pay to at least have a fluent proof reader or editor approve the copy before you run with it.

* The author of this blog makes no pretense to be fluent in Welsh. Comments and corrections welcome.


Taux de Foisonnement – Expansion Rate

by Peter Corless

“Expansion rate.” Like the titular creature of the 1958 movie The Blob, it can be a horrific monster.

Anyone who has had to design web or mobile UI/UX for various international audiences has run into it. So has anyone who has made different language versions of the same web banner ad or multi-page brochure or book layout. Our own CEO, Michel Lopez, has written in the past on the subject. His observations bear revisiting.

Expansion Rate, Defined

When you go from one language to another, and you end up with more characters, even more words, that is the expansion rate for translation. Across the industry, it is usually defined by the percentage of additional words needed to describe something from one language in a different language. For example, when translating from (US) English-to-French, you will typically find that the text “expands” by about 15% – 20%.

Conversely, when you would go from French-to-English, you would see a contraction rate of 13% – 17%. That is usually less of an issue, but still may set visual or editorial professionals needing to deal with truncated translated content twitching.

To give a hat-tip to our industry colleagues over at Kwintessential in the UK, they have a rather handy table of what you can expect for expansion or contraction rates between twenty languages (and regional variations).

Character Expansion

It is not just word-based expansion that authors and designers need to worry about. As the W3C notes, in their document for translation expansion for the web, Chinese and Korean characters often require about twice the visual space (width) per character as Latin-alphabet languages.

Look at the very title of this article: “Taux de Foisonnement – Expansion Rate.” In English, it is “Expansion Rate” (14 characters total, including one space). But in French we have «Taux de Foisonnement» (20 characters, including two spaces). That’s an expansion rate (and converse contraction rate) of +43% (–30%).

It Even Happens in Same-Language Localization

Even localizing text from US English to UK English will see some expansion, often in terms of character-count. For example, the typical alteration of “armor” to “armour” and “color” to “colour.” But there are exceptions. While American “soccer” (6 characters) is shorter than British “football” (8 characters), the American English word “elevator” (8 characters) is twice as long as the British “lift.”

It can cut both ways, idiomatically. The abbreviated American vernacular “gas” is half-as-long as the British “petrol,” but the full word “gasoline” is longer. Which you use may depend on the context of the content, such as whether its use is conversational or technical.

How a Translator Can Help

Don’t just shrink the font size to cram everything into the space available!

When visual real-estate and reader comprehension matters, high-quality human translators and editors can make sophisticated and subtle decisions on how to fit translated content for space. Their toolkit is their linguistic expertise: making idiomatic language and phrase substitutions, removing excessive or repetitive verbiage, using simpler (but still precise) words to convey the same meaning, rewording passages entirely (if need be), and any other necessary modifications.

For these issue aren’t solved by the science, but by the art of translation.

Have you ever run into issues of expansion or contraction in your own translation projects? Have you ever had a project expand like The Blob past the boundaries of your dialogue boxes and page counts? How did it affect the quality of the end products? If you have had past experience, or a current translation project requiring the keen eyes of our linguists, let us know! Email us at projects@e2f.com.


Hitler’s “Mein Kampf” in French Translation

by Peter Corless

Now, more than 90 years since it was first written, and 70 years since the death of Adolf Hitler, Mein Kampf has entered the public domain. The publication and translation of his works has created ethical dilemmas on multiple levels since it first appeared in print to the present day.

The Conversation is an academic and research journalism project has an aim to “rebuild trust in journalism.” It makes an ideal platform to weigh in on the sobering, controversial topic of Hitler’s blueprint for conquest. As Americans enjoyed commemorating their liberation through their July 4th Independence Day celebration, Manu Braganca, Research Fellow at Queen’s University Belfast, published a historical perspective on the 90-year-old book that would lead to the subjugation and decimation of tens of millions and war that spanned the globe.

The article bears reading in full. Because it isn’t about the contents of the book itself, but the context around how Mein Kampf was translated into French, both before and after the Second World War. First, how it was used by an opportunistic French publisher as a tool of profit, which was then suppressed by Hitler himself until he retooled and re-released it for Nazi propaganda purposes.

International translations have long been a source of misappropriated intellectual property and misconstrued meanings. Here, the misappropriation was committed by the original translation: an unauthorized, though rather faithful-to-the-source edition released in 1934 by Nouvelles Éditions Latines. The 1938 re-release, published by Fayard, was a purposeful misconstrual: expurgated and bowdlerized under Hitler’s direction specifically to keep the French from being alarmed by the original fiery and bellicose German. Even the title was changed to Ma Doctrine (My Doctrine) to sound less threatening. Produced just a year before the dawn of the Second World War, the same year as the fateful Munich Agreement, it was a purposeful attempt to sow misunderstanding and allay suspicions for political ends.

Braganca details the fate of the book after the Second World War on to the present day, and presents multiple sides of the debate. Should the book be blacklisted and banned outright? Published with additional front matter and addendum to provide historical context? In a modern world that is increasingly polarized and radicalized, and in a world wide web where historical texts can be found online for free, such controversy has been highly charged. One thing is for certain: debate over the book in France and elsewhere will not end any time soon.

What are your thoughts? Let us know at projects@e2f.com.


University of Hawai’i at Mānoa dives into language, history & science via online newspaper archives

by Peter Corless

Image of two fishermen with outrigger canoes, Waikiki beach, Oahu Island, c. 1922. Source: Wikimedia.

The Institute of Hawaiian Language Research and Translation (Ke Keʻena Noiʻi A Unuhi ʻŌlelo) at the University of Hawai’i at Mānoa has launched a new research program, working with the archives of Hawaiian-language newspapers.

The program, entitled Ka Wā Ma Mua, Ka Wā Ma Hope (Using the Past to Inform the Future: English Translation of Hawaiian Language Newspaper Accounts of Unusual Weather Events), part of the Hawaiian Language Newspaper Translation Project, was funded in 2015 through UH Sea Grant via the NOAA Preserve America Initiative program.

As reported by Hawaiian Public Radio, “The largest source comes from more than 100 different Hawaiian language newspapers published until the late 1940s. Of those, only a small portion have been translated into English, and remain an untapped source of information into Hawai’i’s past.” The last Hawaiian-language newspaper, Ka Hoku o Hawaii, was shuttered in 1948, as noted in a paper by UH Mānoa’s Richard Keao Nesmith.

Hawaiian stands at a crossroads. As noted on Ethnologue, “Young speakers are being trained in immersion courses and also very old speakers exist, but relatively few adult and middle-aged speakers, which results in lack of communication situations for active use.” This younger population is comprised of tens of thousands schooled in Hawaiian studies that date back to 1980, and immersion programs begun in 1986. The number of traditional native speakers has dwindled with time, to just a few thousand, with many now being in their 70s or 80s.

In his paper, Nesmith dubbed the former community the Neo Hawaiian Community, (NEO), and the latter the Traditional (TRAD) community. The differences aren’t just in the vocabulary, or the grammar or dialect. They have to do with the corpus of culture. Nesmith observed, “While most Hawaiians of my grandparents’ generation were fluent native speakers who could recall accounts of Hawaiian heroes, genealogies, chants, and old songs, most of my mother’s generation know virtually nothing of these things, and thus cannot pass them on to their children, grandchildren, and great-grandchildren, even if they want to.”

UH’s collection of newspapers will not restore such collective mythology, but it does bridge the gap between earlier generations and the current ones.

Such native-language news sources are fixed in the crossroads of time. The first Hawaiian-language newspapers were published in 1834 in what was then the independent Kingdom of Hawai’i by missionaries working with the local population. Local literacy and linguistic study of Hawaiian flourished thereafter to the end of the 19th Century.

However, education in Hawaiian was formally banned after the overthrow of the Kingdom, by the 1896 Laws of the Republic of Hawai’i, in a section of law known as Act 57. Throughout the first half of the 20th Century, until the shuttering of the last Hawai’ian newspaper in 1948, literacy in Hawai’ian receded like a wave, stigmatized by the successive dominant English-language American-culture governments of the archipelago.

These newspapers thus represent a major corpus of culture prior to the modern era, before Hawaiian studies were legally and formally resumed. The NOAA grant is specifically geared to using these primary sources to understand historical perspectives on the Hawaiian environment and fishing practices, both local/indigenous, as well as those introduced to the Hawaiian community. The website displays the original article side-by-side with an English-language translation.

Weather events can be searched based on keyword categories, such as “storm,” “calm,” “lightning” or “thunder,” “cold” or “hot.” There are also ways to search articles related to geology, celestial events, and, also of primary interest to the NOAA, fishing.

For example, an article from 28 April 1838, entitled Na Nai Ma Kahana is translated as “The Dolphins at Kahana.” Its 29 typeset lines, which originally appeared in the newspaper Ke Kuma Hawaii, describe how the villagers of Makaua, from the strongest man to the weakest child, paddled canoes out into the ocean at Kahana Bay off the island of Oahu one afternoon to hunt dolphins. The catch of hundreds of the creatures fed the people, as well as their pigs and dogs, and also provided copious amounts of oil for lamps. The reporter signed their eyewitness account simply, “By me, NAILI.”

From such eyewitness accounts, researchers of different disciplines can identify a variety of information. Not least of which would be anecdotal evidence for marine biologists of sea life depletion over time. For perspective, in 1838, the villagers of Makaua killed over 200 dolphins in one afternoon alone. Today, there are three types of dolphins that inhabit the Hawaiian Islands. Of those, there exist only approximately 3,350 spinner dolphins, 3,200 bottlenose dolphins, and 10,260 spotted dolphins, for a total estimated population of 16,810. That afternoon’s catch, if it were held today, would represent more than 1% of the remaining dolphin population of the islands.

Have you ever had a translation project that was more than what it seemed at first? One that expanded profoundly beyond pure linguistics to cultural, scientific, historical, even political dimensions? Maybe you even have one facing you right now. We’d love to hear more! Send us your thoughts, feedback, and current objectives to projects@e2f.com.



Brexit and its Impacts to Localization

by Peter Corless

The vote last night for the Brexit has already had immediate and momentous global financial impact, with markets rocked around the globe from the New World to Asia. The pound plunged to its lowest levels in over 30 years. An estimated $2 trillion to $3 trillion in value was immediately erased in financial markets around the world, and the results led to the resignation of the Prime Minister, David Cameron.

The impacts of this vote will take years and decades to resolve, and will affect the British Isles and the continent in a myriad of ways, some more subtle and some more overt.

Many in the UK voted “Leave” to expressly avoid EU red tape. The EU tried to wiggle a bit and reduce the more odious requirements, but to little effect. For instance, in early June the EU eliminated some translation and localization requirements, such as lifting the requirement for certified translations of certain personal documents. Yet such minor steps fell far short of what British voters seemed to demand.

On the other side, in the “Remain” camp, just days before the Brexit vote, the collective Welsh, Gaelic, Irish, and Cornish language communities issued a joint letter, warning that a vote to leave could be “potentially disastrous” for them. Under the EU, such minority languages had particular acceptance and protection under law. [For example, read our article from earlier this year about Gaeilge being made one of the official languages of the EU.]

The joint statement warned of the consequences of removing such protections. Their language was stark, stating that they would “be at the mercy of governments that have shown neither the interest nor the desire to protect and promote the rights of speakers of our nations and regions’ languages, and have throughout much of our shared history conducted aggressive language policies designed to eradicate our languages.”

Regional concerns and considerations beyond linguistics also highly affected local results. For instance, isolated Gibraltar, which has a land border with Spain, and commonly speaks a vernacular cobbling of English and Spanish known as Llanito, voted nearly 96% to Remain. Scotland voted 62% to Remain, and every single Scottish county voted majority Remain. Northern Ireland, which shares a land border with the EU-member Republic of Ireland, also voted majority Remain with 55.8%.

On the other hand, Wales was nearly evenly split, but had Leave barely ahead at 52.5%. Cornwall also voted majority Leave, with 56.5% of the vote.

In the aftermath, there remains great uncertainty, with calls for a new independence referendum in Scotland, plus Northern Ireland again raising the possibility of reuniting with the Republic of Ireland. Both efforts are directed squarely and specifically this time to regain EU membership.

As an industry, most UK localization and translation professionals wished to Remain, with an ATC poll showing support among 90% of its members. The British linguistics industry accounts for over 12,000 jobs and £1 billion in business annually. The statement from ATC warned “a departure from the EU will cause a significant deterioration in current levels of business they conduct with enterprises based in the member states.” The Brexit can lead to EU citizens currently in the UK having to return to the Continent, or might raise barriers to doing business between British Language Service Providers (LSPs) and clients in the EU, and the vice-verse of these situations.

Even adjacent industries, such as marketing and design, will be affected. In an article in The Drum, Julia Beardwood, of Beardwood and Co., seemed almost sanguine, “Britain is dominant in the world of creativity and Brexit won’t change that… From a branding perspective, global brands are already recognizing the need for more localization. Less ubiquitous campaigns and more country and regional pride is another trend we will see more of.”

Less sanguine was Karin Drakenberg, founder of StrawberryFrog, who said, “The exit of the U.K. from the EU will mean an immediate change of mood among European brands about working with London based agencies…The other issue is that uncertainty will immediately create a state of fear and that means less money for new ideas and innovations. Quite frankly, people will stop investing in new things until there is clarity about this massive change.”

Update: On 27 June, it was observed by a member of the European Parliament that if the UK continues with a Brexit, English would no longer be an official language of the EU! It is possible there might be a way to have another country (such as Ireland) support English as an official language of the EU in place of the UK, yet this is surely a case of broad-ranging unintended consequences of a Brexit.

What is your opinion? How will Brexit affect your own localization and translation decision-making? We’d love to hear your thoughts and concerns. Email us at projects@e2f.com.


Translating the Magic Kingdom

by Peter Corless

When you have one of the most recognizable brands in the world, it may be an alluring temptation to simply presume your global fans know you already, and will understand your meaning. Not so. For a company like Disney, with so many sub-brands of movies and characters established throughout its 92-year-long history, they understand many are more, or less, recognizable in some markets than others. And they take great pains to ensure that each time they make a major step in a market, they’ve taken local language and culture into account.

For instance, with the opening of Shanghai Disneyland this month, every aspect of translation and localization required analysis and, if necessary, rethinking. A Wall Street Journal feature article covered a few of the details. First thing to keep in mind: Disney did not wait until the last minute to do localization. They have been carefully planning their entry for the past six years. They sweated the details, and they made sure to double check by testing it with fans long before the park officially opened on 16 June.

One emblematic example: the iconic Dumbo, who first flew onto American movie screens prior to the bombing of Pearl Harbor in October 1941, was relatively unknown to Chinese Disney fans. For the Shanghai theme park, though he is listed on the English language version of their website as “Dumbo,” for local markets he was translated into Chinese in 2016 simply as “Little Flying Elephant.”

When it comes to culture, the world is not flat. Disney cannot take our culture and export it to China.

So said Bob Iger, Chairman and CEO of Disney in 2013, as quoted by China Daily. He also said at the time, “There will be entertainment and show elements that will be very Chinese in nature, performed by Chinese, and designed, directed and created by artists from China.” Even the menu is localized, as a “giddy” Iger recently enumerated to The Hollywood Reporter, “We’ve got a Szechuan chicken burger, a Peking duck pizza shaped in a Mickey head… There’s something called siu mai dumplings that are Shanghainese — they’re tremendous!”

This isn’t Disney’s first foray into China either. Their Hong Kong theme park opened over a decade earlier in 2005, and lessons learned there will certainly apply. For instance, in Hong Kong, the “Haunted Mansion” changed its name to “Mystic Manor” and removed the ghosts, since Chinese audiences did not view them as particularly fun. Disney likewise has extensive history in localizing its brands, such as the Chinese version of High School Musical, created in 2009.

Though as that foray shows, sometimes Disney might even take things a step too far, such as wanting to replace a basketball scene with kung fu. (Note: China loves basketball, and has since it was introduced to the country in 1895.)

Thus, the 2016 opening of Shanghai Disneyland is the culmination of decades of efforts involving a very large Disney cast, including Iger, who has been involved with negotiations dating back to 1998.

What are your thoughts for your own company, and its brands? Is localization something you plan out far in advance, or is it something left until late (possibly too late) in the process? Conversely, have you ever over-thought a localization process? We’d love to hear. Write to us at projects@e2f.com.