Wednesday, 24 October 2018

Facts About Sweden - the Largest Country in the Nordic Region


Credit: Baltic Media

The largest country in the Nordic Region is also the one with the biggest population. For many, Sweden is synonymous with the production of high-quality cars, iron, and steel.
Inland lakes and large rivers cover almost 10% of Sweden's land mass. Despite the country's enormous coniferous forests, it still has 27,000 km2 of arable land.
Sweden is also the Nordic Region's most populous country, with around 9.1 million inhabitants, almost two million of whom live in the Stockholm area. The northern parts of the country are sparsely populated.
Sweden is a constitutional monarchy. King Carl XVI Gustaf has no real political power, and the parliament, Riksdagen, is the country's highest authority.
Sweden is a member of the EU but has retained the crown as its currency. The country is not a member of NATO.
Sweden is a major exporter of goods and services. The biggest exports are electronics, engineering, cars, paper, iron and steel.
Per capita GDP is € 28,200 (2006).
Total area (1): 447,420 km2
Lakes and streams: 40,080 km2
Arable land and gardens: 26,080 km2
Forests and plantations: 282,760 km2
Largest lake: Vänern 5,648 km2
Highest point: Kebnekaise 2,106 m
Coast line mainland: 11,530 km
National borders: 2,205 km (border to Finland: 586 km, to Norway 1,619 km)
Icecap and glaciers: 283 km2
Population February 2017: 10 014 873 inh.
Population density 01.01.14: 23.7 pop./km2
Population, capital 2016: 2 163 042 inh Stockholm (municipality) (2)
National day: 6th June (Swedish Flag's Day)
Form of government: Constitutional monarchy
Parliament: Riksdag (349 seats)
Membership of the EU: From 01.01.95
Membership of NATO: No
Head of State (as of September 2009): Kong Carl XVI Gustaf
Head of government (as of October 2014): Prime Minister [Stefan Löfven]
Currency: Swedish crown (SEK)
Official website: www.sweden.se
1) Incl. the largest lakes, but excl. the area from the coastline to the territorial boundary, 81,502 km2
2) Stockholm, Upplands Väsby, Vallentuna, Österåker, Värmdö, Järfälla, Ekerö, Huddinge, Botkyrka, Salem, Haninge, Tyresö, Upplands-Bro, Nykvarn, Täby, Danderyd, Sollentuna, Södertälje, Nacka, Sundbyberg, Solna, Lidingö, Vaxholm, Norrtälje, Sigtuna and Nynäshamn
Source: Norden

Saturday, 1 September 2018

We can help you capture new Nordic and Baltic markets. Here is a piece of advice on how to start

Website translation services for ecommerce
Credit: StoryBlocks. Norway

Website translations
Exports of products and services to other countries is what both you and your country need. If you want to capture new markets, having your website or online store only in English is not enough.

According to market research, an online store or website in client's native language has a significant impact on buyer's choice. Roughly 75% online store customers give preference to the seller who provides product and service description in customer's native language.
So if you want to expand your product or service market, start with website localization in the language used on your potential market. However, do not forget to ensure customer service in their native language.
Saves money
  
We have over 25 years of experience in website localization in Nordic and Baltic languages. Our clients will also save money as we use CAT (Computer Assisted Translation) tools that create a translation memory and terminology database so you will not pay twice for phrases that repeat 100% and will pay less for those somewhat similar to the translated ones. Terminology database will also ensure terminology consistency and search engine optimization.
Our translators are human beings instead of machines as machine translation such as Google cannot understand the context, idioms, jokes and other nuances, therefore, nothing has been able to beat professional translators so far. The leading global and local scale companies still localize their content only with human translation.
We recently localized Swedish hairdresser equipment and product wholesaler website in all Scandinavian languages - Headbrands.seHeadbrands.fiHeadbrands.dkHeadbrands.no. Their turnover saw increases of growth!
Find more information on website translation and our work flow on our website.

Monday, 13 August 2018

English's Days as the World's Top Global Language May be Numbered. Can English Remain the 'World's Favourite' Language?



English is spoken by hundreds of millions of people worldwide, but do the development of translation technology and "hybrid" languages threaten its status?
Which country boasts the most English speakers, or people learning to speak English?
The answer is China.
According to a study published by Cambridge University Press, up to 350 million people there have at least some knowledge of English - and at least another 100 million in India.
There are probably more people in China who speak English as a second language than there are Americans who speak it as their first. (A fifth of Americans speak a language other than English in their own homes.)
But for how much longer will English qualify as the "world's favourite language"? The World Economic Forum estimates about 1.5 billion people around the world speak it - but fewer than 400 million have it as their first language.
Of course, there is more than one English, even in England. In the historic port city of Portsmouth, for example, the regional dialect - Pompey - is still very much in use, despite the challenges from new forms of online English and American English.
English is the world's favorite lingua franca - the language people are most likely to turn to when they don't share a first language. Imagine, for example, a Chinese speaker who speaks no French in conversation with a French speaker who speaks no Chinese. The chances are that they would use English.
Five years ago, perhaps. But not anymore. Thanks to advances in computer translation and voice-recognition technology, they can each speak their own language, and hear what their interlocutor is saying, machine-translated in real time.
So English's days as the world's top global language may be numbered. To put it at its most dramatic: the computers are coming, and they are winning.
You are probably reading this in English, the language in which I wrote it. But with a couple of clicks on your computer or taps on your tablet, you could just as easily be reading it in German or Japanese. So why bother to learn English if computers can now do all the hard work for you?
At present, if you want to do business internationally, or play the latest video games, or listen to the latest popular music, you're going to have a difficult time if you don't speak any English. But things are changing fast.
In California, Wonkyum Lee, a South Korean computer scientist for Gridspace, is helping to develop translation and voice-recognition technology that will be so good that when you call a customer service helpline, you won't know whether you're talking to a human or a computer.
Christopher Manning, professor of machine learning, linguistics and computer science at Stanford University, insists there is no reason why, in the very near future, computer translation technology can't be as good as, or better than, human translators.
But this is not the only challenge English is facing. Because so many people speak it as their second or third language, hybrid forms are spreading, combining elements of "standard" English with vernacular languages. In India alone, you can find Hinglish (Hindi-English), Benglish (Bengali-English) and Tanglish (Tamil-English).
In the US, many Hispanic Americans, with their roots in Central and South America, speak Spanglish, combining elements from English and the language of their parents and grandparents.

Language is more than a means of communication. It is also an expression of identity - telling us something about a person's sense of who they are. 

The San Francisco poet Josiah Luis Alderete, who writes in Spanglish, calls it the "language of resistance", a way for Hispanic Americans to hold on to - and express pride in - their heritage, even if they were born and brought up in the US.
English owes its global dominance to being the language of what until recently were two of the world's most powerful nations: the US and the UK. But now, especially with the rise of China as an economic superpower, the language is being challenged.
Computerised translation technology, the spread of hybrid languages, the rise of China - all pose real challenges. But I continue to count myself immensely fortunate to have been born in a country where I can cherish and call my own the language of Chaucer, Shakespeare, Milton and Dickens, even though the language I call English is very different from theirs.
Presentational grey line
Read more: By Robin Lustig Presenter, The Future of English, BBC World Service 

Friday, 10 August 2018

- Iceland 24 - Iceland Travel and Info Guide : History of the Icelandic Language

- Iceland 24 - Iceland Travel and Info Guide : History of the Icelandic Language: We’ll be the first ones to admit it. The Icelandic language is not one of the bigger, more popular languages like English, Spanish or French...

Credit: Baltic Media

What makes the Icelandic language special
Iceland is an island in the North Atlantic Ocean situated just below the Arctic Circle. When you look at Iceland on a map, it seems pretty lonely. But this isolation did have a linguistic benefit. Iceland’s location prevented its language from being too diluted or influenced by languages of nearby countries. As a result, Icelandic is the North Germanic language that most closely resembles the Old Norse spoken over a thousand years ago. The Viking Sagas and Eddas written around 800 years ago can still be read today because the language really has not changed that much. Not even English can say that!

Sunday, 22 July 2018

Nobody should ever trust a translation produced by a machine. Only a human translator can verify the machine did a good or a bad job… and fix it



In the past years, I’ve been evaluating machine translation output extensively. I started to do this because I was curious, and I ended up doing this professionally. As this evaluation work was for various NLP scientists and translation agencies,  I cannot dwell on details, but what I learned in the various tests is this:
  • The machine translation output has overall improved a lot. 5 years ago a test document with 1000 sentences contained more than 700 sentences where a translator was taking the risk to waste time editing the machine translated text. Today less than 100 sentences are too bad for editing. “Too bad for editing” means that the sentence the translator creates for delivery is too different from the machine translated text: too many words had to be dropped, added, moved or edited.
What changed in these 5 years? The test set was almost the same: the same source sentences, but different engines. The Systran, Google and Microsoft engines became neural MT systems, and I added Amazon and DeepL to the translation set. (A pity I could not add LILT to the test set — I cannot use adaptive MT in my measuring tools. But I’m sure LILT has improved at the same pace as the others. Maybe even more…) In the same test set, there were also high fuzzy matches from translation memories. Those I never changed as I used them for validating the data.

The sentences in the test set were quite long (15 to 45 words, or up to 300 characters), and the syntax of some sentences was quite complex. Clearly, 5 years ago, MT system could not handle well those long sentences. Today the quality of the machine translated long sentences is much, much higher.
Another observation: translators were asked to label sentences before editing — we asked them if they believe the pre-translated text was retrieved from a translation memory or if it was produced by an MT engine. 5 years ago most translators got it right, because “machine translated sentences contain more mistakes all over the place“.  This year, they could also tell what was MT, but the reason was very different: “the edits I had to do on TM output were mostly fixes of mistakes MT systems don’t make“.

Interesting evolution.
Of course,
  • there are still big differences between the language pairs. For some target languages, I really doubt MT will ever be good enough to be used by professional translators. But for a lot of target languages, it is clear: machine translation is a solid tool that can help to translate more in less time.
  • there are still a lot of documents that will never be fit for MT pre-translation. It has just become easier to spot them now.
  • nobody should ever trust a translation produced by a machine. Only a human translator can verify the machine did a good or a bad job… and fix it.
So I was wondering: why do we keep on using the term “post-editing”? Maybe we should stop using this term, and just use “editing” or why not just use “translating”? This is what translators do, after all…



Source: Gert Van Assche  https://theopenmic.co/post-editing/

Tuesday, 19 June 2018

Global Market for Outsourced Translation and Interpreting Services and Technology to Reach US$46.52 Billion in 2018



Common Sense Advisory's 14th annual comprehensive study of the language industry shows growth continues due to content digitization, personalized customer service, and business globalization.
The global market for outsourced language services and technology will reach US$46.52 billion in 2018, according to a primary quantitative study by independent market research firm Common Sense Advisory (CSA Research). The firm surveyed providers from around the world to collect actual reported revenue for 2016, 2017, and expected revenue for 2018. CSA Research details its findings in the 14th annual global industry report, "The Language Services Market: 2018," the only comprehensive global analysis of private and publicly-traded language services and technology companies.
As organizations both large and small make their products and services available in more languages, the firm predicts that the language services industry will continue to grow and that the market will increase to US$56.18 billion by 2021.

The firm found that the demand for language services and supporting technologies continues and is growing at an annual rate of 7.99%, representing an increase over last year's rate of 6.97%. 

Sixty-four percent of surveyed language services providers (LSPs) said revenue was up over the previous year. Factors driving this demand include content digitization, personalized customer service, and business globalization.
  • Revenues, rankings, and locations of the 195 largest LSPs in the world
  • Regional rankings of the largest LSPs in AfricaAsia-PacificEastern EuropeLatin AmericaNorth AmericaNorthern EuropeSouthern Europe, and Western Europe
  • Trends in automation and spoken-language technologies including the impact of artificial intelligence on project management automation
  • Breakdown of the market by translation, interpreting, localization and engineering, project management, and more
  • Breakdown of the market for technology sold by LSPs and technology providers with estimates for translation management, translation memory, terminology, machine translation, interpreting management, and other software
Common Sense Advisory (CSA Research) is the premier market research firm specializing in the language services and technology industry. It provides primary data and insight to assist companies with planning, brand strategy, innovation, competitive positioning, and better understanding of global markets.


Included in report series are the largest 195 language service providers globally, as well as by region, all of which offer language and localization services to enable enterprises to expand global reach and to respond to domestic needs. The top 10 largest commercially-focused language services companies worldwide, listed according to 2017 revenues, are: TransPerfect; Lionbridge; LanguageLine Solutions; RWS Holdings plc; translate plus; SDL; Hogarth Worldwide Limited; Welocalize; Amplexor International; and Keywords Studios.
CSA Research's structured and documented market research methodologies ensure comprehensive and independent data-driven research for LSPs, technology vendors, global enterprises, and investors. Primary data and insight in CSA Research's 2018 independent study include:
"The Language Services Market: 2018" is available to CSA Research members. The list of the largest LSPs based on 2017 revenues is open-access and available here.
About Common Sense Advisory
Contact: Simona Bertozzi, 978-275-0500, 197258@email4pr.com
SOURCE: PRNewswire.com  
Related Links

Sunday, 6 May 2018

Scandinavian Languages: Are the Three Neighbouring Languages Becoming Strangers*?


Translation Nordic languages
Credit: StoryBlocks


The languages spoken in Scandinavia are called North Germanic languages and include Danish, Swedish, Norwegian, Icelandic, and Faroese. They are subdivided into East- (Danish, Swedish) and West-Scandinavian (Norwegian, Icelandic) languages. Finnish, being completely different, belongs to the Finno-Ugric language family.
Danish, Swedish and Norwegian are all very similar, and it is common for people from all three countries to be able to read the two other without too much difficulty. Understanding the spoken languages, however, can present more difficulties, especially for Swedes and Norwegians who have had little exposure to spoken Danish.
Icelandic and Faroese do have some words in common with the three other Scandinavian languages, but it is not common for Scandinavians to be able to understand Icelandic and Faroese, except for certain Norwegians who have a similar dialect (Norwegian nynorsk).

Norwegian is to Danish as Valenciano is to Castellano
In terms of vocabulary, the most similar languages are Danish and Norwegian, which is possibly due to Norway once being under Danish rule. The two languages differ about as much from each other as Castellano does from Valenciano. The main difference lies in the spelling of and pronunciation of words – the words often being the same words and having pretty much the same meaning, just spelled slightly differently. In some cases, however, a certain word will be used in Norwegian and another in Danish in the same way as English for example has ‘lorry’ and ‘truck’.
Whereas written Danish and Norwegian (Norwegian bokmål) is very similar, the written Swedish language contains some words that a Danish and Norwegian person cannot possibly understand unless they know them beforehand.

“Norwegian is Danish spoken in Swedish
When it comes to pronunciation, however, Swedish and Norwegian are very close. For a Dane and a Norwegian, it can sometimes be quite hard to communicate, as Norwegians tend to ‘sing’, wheras we Danes ‘talk as if we have a potato in our mouth’. Swedes also ‘sing’, but depending on the region, some Swedish speaking people are easier to understand for Danes than Norwegians because they don’t ‘sing’.
In conclusion, this gives us the following equations:
Norwegian + phonology – vocabulary = Swedish
Norwegian – phonology + vocabulary = Danish
Thus, it has been said that “Norwegian is Danish with Swedish pronunciation”. While this is not completely accurate, there is some truth in the statement.

Are the three neighbouring languages becoming strangers?
Despite the similar languages, Scandinavians sometimes end up speaking English between themselves often due to the dialects existing in the Scandinavian countries and the effects of globalisation. However, making an effort to try and understand each other is only a matter of practice – in the same way as when an American person tries to understand a Scottish person.
If you were to learn one of the Scandinavian languages, which one would you go for?

*Written by Claus Skovbjerg, MA, stagiaire communicateur at TermCoord
Read more: Termcoord.eu 


Thursday, 5 April 2018

Website Translation: Creating Content in Multiple Languages

Credit: storyblocks

There are a few common scenarios when creating content in multiple languages. Determining which of these matches your situation is key to making the right decision when building your site and tackling your website SEO.
The three main scenarios we see when building multi-language websites are:
  • Multiple languages serving the same country.
  • Multiple languages serving no specific country.
  • Multiple languages serving multiple countries.
Let’s look at each of these in a little more detail so you can understand what is the right choice for you.
Multiple languages serving the same country
Canada is a good example since it is one country with two official languages, English and French.
 Here we could have a single website serving a single country with multiple languages. In this case, we would want to use a .ca country code Top-Level Domain (ccTLD) for Canada to automatically geo-locate the site and then have content in English and French to target French- and English-language queries.

Sometimes it is easier to consider what can go wrong here:
  • Is the English language version of the United Kingdom (UK)?
  • Is the English language version for the United States (US)?
  • For Australia? Or all the above?
  • Is the French version for France?
To ensure the search engine understands your site, geo-targeting and language targeting multilingual SEO tactics should include:
  • ccTLD for the country being served to benefit from default geo-targeting.
  • Single website with language-specific content in subfolders English and French: /en/ & /fr/.
  • Site hosted in the country that is being targeted.
  • Hreflang tags specifying language and country.
  • Links from relevant specific language websites.
With all of these steps followed, a search engine has all the pointers needed to know that this content is for English and French language speakers in Canada.
Multiple languages serving no specific country
Here we have a situation where we are targeting users based predominantly on their language.
We are not concerned if an English speaker is in the UK, the US or Australia or any other English language speaking location (small differences in spelling aside).
We don’t care if this is an Englishman in a country that speaks another language. As additional languages are added, they target speakers of that language around the world with no geographical bias.
Imagine a company that provides a software solution around the world. This business will want to have content in each language and have search engine users find the correct language version of the content.
So, an English speaking visitor in the UK, the US, Canada or Australia would all get the same content. A French speaker in France or Canada would also get the same French content.
Options here are a little more diverse. This is where considerations from the real-world and business operations become crucial in making the right decision (as discussed in more detail in my international SEO guide).
The tactic we recommend in this scenario is a single site with the following multi-language SEO tactics in place to support the desired ranking goals:
  • A generic TLD such as a .com that can target multiple countries.
  • Single website with language-specific content in subfolders (e.g., /en/, /fr/, /de/).
  • Site hosted in primary market with an international content delivery network.
  • Hreflang tags with language-only specified (not location).
  • Links from relevant specific language websites.
As the world gets smaller and subscription-based software solutions become ever more popular, this kind of setup is a simple way to target multiple languages across the globe.
Multiple languages serving multiple countries
This is where things can get a little more complicated because we may have multiple versions of the same language with nearly duplicate content, so technical configuration needs to be 100 percent accurate.
We may have a site in English and French, and we may have an English language section for each of the UK, the US, Australia, and Canada, along with a French page for France and Canada.
  • www.example.com/ — US, English (default).
  • www.example.com/gb/en/ — UK, English.
  • www.example.com/au/en/ — Australia, English.
  • www.example.com/ca/en/ — Canada, English.
  • www.example.com/ca/fr/ — Canada, French.
  • www.example.com/fr/fr/ — France, French.
This is fairly basic: two languages and five locations. We have seen this get a lot more complicated, and if it confuses you, then the odds of tripping up a search engine are amplified!
Get this wrong and your rankings go down the international SEO tube.
Tactics here for a single site include:
  • A generic TLD such as a .com that can target multiple countries.
  • Default location and language (US English in this example).
  • Country-specific subfolders (gb/, au/, ca/, fr/).
  • Language-specific subfolders below the country-specific subfolders (gb/en/, ca/fr/).
  • Site hosted in primary market with an international content delivery network.
  • Hreflang tags with language and location specified.
  • Relevant links from location- and language-specific websites.
This is a straightforward way to achieve the targeting of multiple languages in multiple locations.
Read more: 

SEO for multi-language websites:How to speak your customers’ language




Thursday, 1 March 2018

Andy Martin*: Google Translate will never outsmart the human mind – and this is why

Translation by Humans and Machine translation, translation google translator
Credit: StoryBlocks
... Google is often adequate, and in so many languages too, but only in the way of a particularly uninspired apprentice translator. I once picked a book off a shelf in a bookshop because I was attracted by its zany title: Whatever. It was only when I saw the name of the author and leafed through it that I realised that it was a book I already knew well in French: Michel Houellebecq’s Extension du domaine de la lutte. After initially saying to myself, ‘What kind of crazy translation is that?!’, I saw that, in fact, it’s a stroke of genius. The original title is deliberately turgid to the point of being interesting (and may, in fact, be a homage to the sociologist Auguste Comte) and the translator had achieved what is surely the only real point of a title, which is to make you pick up the book in the first place. Google’s “Extension of the field of struggle”, while technically permissible, has the opposite effect.

Perhaps it’s obvious that a machine is going to struggle with the resonance and complexity of, say, Victor Hugo or Jean-Paul Sartre. So I thought I’d start with an easy one. One of the first sentences I (like many others, I suspect) can remember learning, probably around the age of 3 or 4, before even going to school, is this: “The cat sat on the mat.” Google Translate suggests: “Le chat s’est assis sur le tapis.” Again, good try Google. But try remembering that 50-odd years from now. You could argue about the tense and even the choice of noun (is “mat” really “tapis”?) But the main point is that Google can’t see that it’s a mnemonic, made up of rhyming monosyllables, and that the best solution is to change the species, which is what French does, in the children’s rhyme, “Il était une souris qui mangeait du riz sur un tapis gris…” (There once was a mouse who was eating rice on a grey carpet…) Now that I can remember. (And it goes on, “Et sa maman lui dit, ce n’est pas gentil de manger du riz sur un tapis gris.”)

A machine translator does nothing but translate. This is how it sees its job. As a form of tautology or equivalence. One set of words is exchanged for another set of words. One code is replaced by another code. But, you will say, isn’t that what translation is? This is what I tell my class: if you want to be a good translator, don’t translate. Only bad translators translate. You have to live it. If you want to translate George Sand or Flaubert or Tolstoy, for the duration of that translation, you have to be George Sand, you have to be Flaubert, but reborn, as if they really spoke English, now.

There is, at the core of the translation process, a mystery, an almost mystic transcendence. There is no direct equivalence of one language to another. It’s not just that certain words (eg hygge in Danish) cannot be satisfactorily translated: none of them can. This is what happens in a serious translation. You read a sentence. But – and this is the point that Google tends to miss – those squiggles on the page actually represent something other than words, they are not reducible to mere information ones and zeroes. So you convert them into something other than words. Something like ideas, imprecise though that term is. Or feelings. You infuse the words with your own memories, your experiences, your fears and desires, things you have done or seen or fantasised about or heard once in a song on the radio that you will never hear again. The experience of having been born and being doomed to die also get in there. You – for a brief impossible moment – become Tolstoy, and then and only then can you re-express what was said somewhere else in some other time in your own words in your own time. And, inevitably, of course, you still get it wrong.

Translation is like the archetype of all human relations. We never get it quite right when it comes to understanding other people. At the same time, we ought to try. Being human is an advantage when it comes to translating other humans. 

Consider this, for example. Simone de Beauvoir, the philosopher, writes in one of her memoirs, “La religion ne pouvait pas plus pour ma mère que pour moi l’espoir d’un succès posthume.” Google suggests: “Religion could no more for my mother than for me the hope of a posthumous success.” Google here makes no (or little) sense. Mainly because Beauvoir, in her elegant way, is being elliptical and economical. To spell it out (a little laboriously, I admit), she is saying, “The afterlife promised by religion was of no more comfort to my mother than the hope of a posthumous literary success was to me.” It helps if you know Beauvoir was an atheist of course. And also if you have a rough idea of the kind of salvation on offer from religion. If all you can see is a bunch of words, then you’re stuffed.

Or what about this? A touch more obscure and archaic, but not unintelligible. From an essay written several centuries ago by Catherine des Roches (or Kate of the Rocks), who really wanted to be an intellectual rather than have to hang out at balls and parties in pursuit of “courtly love”: “quant à moy, qui n’ay jamais fait aveu d’aucun serviteur, et qui ne pense point meriter que les hommes se doivent asservir pour mon service…” Google proposes the following: “as to me, who has never made a confession of any servant, and who does not think it merits that men ought to enslave me for my service…” This is semantically and grammatically challenged, ie complete garbage. I hate that “me who has”. So clunky. How can “enslave me for my service” ever be right? Here is my version (entering, for a moment, into the mind of a 16th-century #Metoo poet): “And what of me? Not currently in a relationship, nor ever really had one, and don’t really want one either if we have to go through this silly business of a star-struck lover pledging to be my servant.” It’s far from perfect, but at least it sounds like a human being (who may have acquired a FaceBook account).
I admit that every now and then I suspect a mischievous student of putting a Google translation in front of me just to keep me on my toes. 

Machine translators are a relatively recent invention. But machines have been with us for millennia. Technology is a necessary supplement to humanity. But tools and machines are cold dead things, they are essentially inert. We have to switch them on. All of literature and philosophy and expressive language is a protest not against the machine per se, but against people behaving as if they were machines – incapable of making a judgment call in particular circumstances (which circumstances nearly always are). 

The human mind, we are implicitly saying, is something other than a constellation of metal or silicon. Similarly, the good translation is not translation – exchanging a random collection of information for another – it’s more like a form of resurrection. Translation gives you not just the meaning of a text, it gives you the heart and soul of its author. Its secret message is always, “I am not a robot.”

*Andy Martin is the author of ‘Reacher Said Nothing: Lee Child and the Making of Make Me’, and teaches at the University of Cambridge. 

Read more: The Independent