Bible translation organizations in the United States receive more than $500 million in donations per year. So how many Bibles actually get translated? And how much does a Bible translation cost?
Remarkably, the answer to that question is — nobody really knows.
That’s why the Sagamore Institute, an Indiana-based think tank, recently did a study to analyze the cost and use of funds in Bible translation. The study was funded by the Chattanooga-based Maclellan Foundation on behalf of the illumiNations Resource Partners. These are organizations and individuals who fund Bible translation efforts.
The findings include the following:
The average project cost of a “text translation” is $59,302 per year.
The average project cost of a complete written Bible is $937,446.
The “annual expenditure aggregate project cost” is $105 million.
On average, it takes 15.8 years to complete a Bible translation.
According to a statement, “the study distinguished between project costs (project development, accountability, translation tools, and translators) and support costs (local capacity, maintenance of translation tools, research and development, and alliance infrastructure) because support costs are typically underwritten by specific funding partners, rather than outside donors, and include recurring expenses.”
Give a gift of $25 or more to The Roys Report this month, and you will receive a copy of “Untwisting Scriptures: Wolves, Hypocrisy, Sin Leveling and Righteousness” by Rebecca Davis. to donate, click here.
The numbers released by the Sagamore Institute highlight the fact that the vast majority of dollars contributed to Bible translation organizations do not, in fact, go to Bible translation.
For example, the organizations that make up illumiNations took in more than $521 million last year (see chart), and these organizations produce less than 20 complete Bible translations in a year. If it really costs less than $1 million to produce the Bible, as the Sagamore Institute says, that means support and other costs could have topped $500 million.
The Sagamore study also highlights another reality of the Bible translation industry: the practice of money transfers (grants) between Bible translation partners. These transfers mean that simply adding up the revenue of various Bible translation organizations will likely result in double-counting of revenue. That’s one reason the Sagamore study says the amount of money spent on Bible translation is not in the neighborhood of $500 million per year but about $378 million in the 12-month period reviewed. The study excludes “SIL costs outside of those in partnership with Wycliffe USA; grants unrelated to American Bible Society in the United Bible Societies’ International Support Programme, and differences in the treatment of GAAP reconciliation items.”
However, even accepting the Sagamore Institute’s lower number of $378 million, that means more than two-thirds of the money donated annually to Bible translation organizations goes to activities other than Bible translation. The Sagamore study identifies $97 million in “translation support costs” and $109 million in “related Bible ministry costs.” Sagamore also identified $67 million in “activities conducted by translation partners that are unrelated to Bible translation or ministry.”
All of this means that the “fully loaded” cost of a Bible translation is certainly in the millions of dollars, and likely in the tens of millions.
Calvin Edwards has been studying the Bible translation industry for years. He also had questions about “support costs” and why they were not included in the calculations for the cost of a Bible translation.
“What are these?” he asked. “Do they relate directly to Bible translation? If support costs were included, what is the ‘full cost’ of a Bible translation?”
Edwards added, “The reported findings are interesting but not new, and much more information is required. Donors most want to know two things: how is the $500-plus million raised by Bible translators annually used, and how many Bibles are translated for the translation portion of the total? These are simple questions that have answers.”
According to Rob Panos of The Sagamore Institute, the study was based on information self-reported by the Bible translation organizations themselves. He said the numbers were “validated in aggregate through a review of audited financial statements.” He added, “It’s important to understand that this was not a rigorous audit of spending. It is a high-level view of the cost and use of funds in translation.”
This article was originally published at Ministry Watch.
Warren Cole Smith is president of MinistryWatch.com, a donor watchdog group. Prior to that, Smith was Vice President-Mission Advancement for the Colson Center for Christian Worldview.
An advertisement by People For Better Education, Inc., that appeared in the July 7, 1976 edition of the Anchorage Times advocating for the dictionary ban. The ad was notably published after the dictionary had already been banned.
Part of a continuingweekly serieson Alaska history by local historian David Reamer. Have a question about Anchorage or Alaska history or an idea for a future article? Go to the form at the bottom of this story.
Did you know that a dictionary was banned in Anchorage? Every September, every annual banned books week, it comes up again. Every year, the reading community — authors, librarians, teachers, publishers, booksellers, and consumers — come together to discuss ongoing censorship issues and the history of banned books. And the extraordinary decision to ban a dictionary, of all things, in Anchorage is often included in the dialogue. The details are often wrong, and the story is slightly more complicated than how it is usually presented, but the core of the narrative is true. A dictionary really was banned in Anchorage.
In the early 1970s, a group of concerned Anchorage parents led by Marroyce Hall formed a conservative school watchdog organization called the People for Better Education. Their first claim to fame came in 1974 when the group issued an unasked-for report on unruly behavior in local schools. The report cited rampant disciplinary problems, including thrown snowballs, “pop can hockey,” “general pushing and shoving,” and “displays of affection, such as passionate embraces on school grounds.” Given the existence of such unbelievably wicked behavior, the organization demanded schools add a roving security force.
In 1976, the group emerged with a new issue. They discovered a vulgar book present in many area elementary school libraries, the American Heritage Dictionary. The People for Better Education meticulously combed through every entry in the dictionary, a process one member described as “like digging (in) garbage.” When their labors were completed, they had identified 45 objectionable slang definitions.
Many of the offending words were included for their alternative meanings, including ass, ball, bed, knocker, nut, and tail. Per the dictionary, a “ball” can be “any of various rounded movable objects used in sports and games” or refer to “the testicles.” Similarly, “bed” might be a noun or a verb, a piece of furniture, or “to have sexual intercourse with.” The organization also disliked several idioms, including “shack up,” defined as “to live in sexual intimacy with another person, especially for a short duration.”
The Anchorage School Board, to their credit, took the complaint seriously. They empowered a committee of four parents and four staff members to review the text, which concluded that the dictionary should remain available to students. Undeterred, Hall, accompanied by Eileen Kramer, made a presentation to the school board on June 28, 1976, again demanding the removal of the dictionary.
To the surprise of board president Sue Linford, the board voted 4-3 to remove the dictionary. Millet Keller, Tom Kelly, Darlene Chapman, and Vince Casey voted in favor of the ban. Keller, who authored the motion to remove the dictionary, told the Anchorage Daily Times that such “vulgar, slang words” were “better left in the gutter.” Linford, Heather Flynn, and Carolyn Wohlforth voted against the ban. Said Linford, “I’m really at a loss to explain it ... This smacks of things I don’t think we want to encounter.”
The next day, the Anchorage Assembly voted unanimously to send a letter to the school board sharply rebuking their actions, warning that the dictionary ban would “stimulate further censorship pressures.” The letter further stated, “the overriding concern is preservation of individual and academic freedoms. We earnestly request that you reconsider your action without delay.” Several Assembly members also went on the record with more personal comments. Chairman Dave Rose described the ban as “absolutely ludicrous.” Lidia Selkregg declared, “I’m shocked at the board’s action. A dictionary is a most sacred document.” Fred Chiei said, “Now I suppose they’d like to go for the Bible. Lots of good words with dirty meanings there.”
Most of the public comments on the ban criticized the action. School board candidate Pam Siegfried said, “I do not want to have my kid go to a library which has been picked clean by concerned parents. Once somebody starts censoring, where does it stop?” The Anchorage Daily News editorialized, “We can’t imagine any action by a board of education which could go against the American grain more than censoring research tools.” As one resident wrote in a letter to the Daily News, “You don’t stop the use of ‘vulgar’ language by removing one book which contains the definition of four or five slang words but rather by parental guidance and by making it clear to your children that such language is not acceptable to you.”
Around the same time, several members of Anchorage’s Mexican American community, including the Chicano Interservice Association, complained that three books on Mexico in local elementary libraries presented stereotyped images of Mexico. The school board ordered the removal of these books at the same meeting when they banned the American Heritage Dictionary. One of these books, “All Sorts of Things” by Theodore Clymer, includes a story of a Mexican town where everyone is lazy, unintelligent, wears tattered sombreros, and eats tortillas exclusively. A month later, the school board rescinded their ban on two of these books on Mexico, though not on “All Sorts of Things.”
The American Heritage Dictionary remained forbidden. As might have been reasonably expected, the ban increased interest in the book and the supposedly objectionable words, underpinning the ludicrous nature of such censorship. The downtown Book Cache (remember when there were multiple Book Caches in town?) prominently displayed a rack full of copies. A spokesperson noted the store “has gotten more requests and comments” since the ban. She added, “I think people are just curious.”
As Katherine Chamberlain of the American Library Association’s Office of Intellectual Freedom stated in 2010, “Condemning the American Heritage Dictionary for its ‘objectionable language’ in effect condemns the English language itself.” The language existed, whether one considered it vulgar or not. Moreover, children then certainly already knew far more profane words than tame terms like “balls” or “tail.”
In addition, Anchorage of the 1970s was a more openly risqué city than it is today. Prostitutes walked the streets, and there were far more strip clubs, XXX theaters, and adult bookstores. The abundant massage parlors and escort services openly marketed their illicit services, establishments like the Touch N Glow, Foxy Lady, and Sensuous Lady. The People for Better Education might as well have expanded their efforts to include banning the phone book since it included sections for such businesses. The newspapers ran advertisements for XXX features, resulting in movie listing pages that included both Linda Lovelace’s “Deep Throat II” and Disney’s “Bambi.”
The People for Better Education were less successful with their future demands, which included the requested removals of several other books and films. In 1977, they notably challenged the showing of the film “The Lottery” in classes. The film is based on the classic short story by Shirley Jackson, wherein town’s people ritually kill a fellow resident every year solely because it is tradition.
The organization was perhaps more successful as an inspiration for similar censorship movements elsewhere. Amid a general rise in school district book bans across the country, the American Heritage Dictionary was subsequently banned in Cedar Lake, Indiana, Eldon, Missouri, Folsom, California, and Churchill County, Nevada in 1976, 1977, 1982, and 1992, respectively.
Having protected children from selected dictionaries, the Anchorage school board next targeted LGBT teachers. Beginning in the summer of 1977, the board tried to suspend and then fire Government Hill Elementary teacher Michele Lish for the perceived sin of being a lesbian. While Lish had a stellar reputation as an educator, the board might have found a way to dismiss her quickly if they had been willing to endure the due process for such a removal. However, the board had no interest in any public accounting and refused to grant Lish a hearing. By September, the courts had twice blocked the suspension and dismissal given the lack of said hearing. Still, the school board dragged the conflict out until January 1978, when Lish accepted a non-teaching position in the district. Coincidentally, that same month, the Copper River School District board passed a resolution banning LGBT employees.
While the dictionary ban does not seem to have ever been officially lifted, the removal was eventually forgotten or ignored. Today, the Anchorage School District library catalog lists several American Heritage Dictionary editions available throughout the district.
The way history works is that the same scenarios tend to repeat, different in the specifics but consistent at their core. Schools should be open, welcoming institutions. Instead, far too many people far too often expend their energies trying to restrict access to literature, knowledge, and access.
“Board Ousts ‘Vulgar Dictionary.’” Anchorage Times, June 29, 1976, 1, 2.
“Books on Mexico Restored.” Anchorage Daily News, August 11, 1976, 2.
Chamberlain, Katherine. “Spotlight on Censorship—The American Heritage Dictionary.” Intellectual Freedom Blog, Office of Intellectual Freedom, American Library Association, September 28, 2010.
Doherty, Nancy. “Dictionary Falls Over Dirty Words.” Anchorage Daily News, June 30, 1976, 1, 2.
Hunter, Don. “Dictionary Ban Draws Criticism of Assembly.” Anchorage Times, July 2, 1976, 1, 2.
After several years, the doors of the scriptorium and the editorial room of the prestigious Encyclopaedic Sanskrit Dictionary at Pune’s Deccan College Post-Graduate and Research Institute in Pune were opened for students and the general public. The year of completion of this gigantic dictionary project, which commenced in 1948, remains unknown. But the final word count is estimated to touch 20 lakh and would be the world’s largest dictionary of Sanskrit.
The Project
Linguist and Sanskrit Professor SM Katre, founder of India’s oldest Department of Modern Linguistics in Deccan College, conceived this unique project in 1948 and served as the dictionary’s first General Editor. It was later developed by Prof. AM Ghatage. The project is a classic example of painstaking, patient and relentless efforts of the Sanskrit exponents for the last seven decades.
The current torchbearers of the Encyclopaedic Sanskrit Dictionary project is a team of about 22 faculty and researchers of Sanskrit, who are now working towards publishing the 36th volume of the dictionary, consisting of the first alphabet ‘ अ ‘ .
Subscriber Only Stories
Between 1948 – 1973, around 40 scholars read through 1,464 books spread across 62 knowledge disciplines – starting from the Rigveda (approximately 1400 B.C.) to Hāsyārṇava(1850 A.D.) – in search of words that could be added to this unique dictionary.
They covered subjects like the Vedas, Darśana, Sahitya, Dharmaśāstra, Vedānga, Vyakarana, Tantra, Epics, Mathematics, Architecture, Alchemy, Medicine, Veterinary Science, Agriculture, Music, Inscriptions, In-door games, warfare, polity, anthology along with subject-specific dictionaries and lexicons.
In the non-digital era, these scholars noted details of every new word onto small paper reference slips. They mentioned details like the book title, context in which the word was used, its grammatical category (noun/verb etc.), citation, commentary, reference, exact abbreviation, and date of the text. It was signed off by the creator of the slip and its verifier.
It took 25 years for these scholars to complete the word extraction process from around 1,464 books to generate one crore reference slips. All these paper slips have been well preserved, alphabetically, in one of the rarest scriptorium – the soul of the dictionary – inside over 3,057 specially-designed metal drawers. They have also been scanned and preserved digitally.
Word-by-word
Advertisement
While this dictionary contains words in alphabetical order, it follows historic principles in stating the meaning. In addition to the word meaning, the dictionary also provides additional information, references, and context of the respective word used in a particular literature. That is why, it is an encyclopaedic dictionary wherein words have been arranged according to the chronological order of their references appearing in the text.
For example, the word beginning with the letter ‘ अ ‘, like agni, will have all the citations from Sanskrit texts starting with Ṛgveda and the references from the texts following Ṛgveda, chronologically arranged. This helps a reader to understand the historical development of the meaning of the word.
“Sometimes, a word can have anywhere between 20 to 25 meanings as it varies depending on the context of use and books. Once the maximum possible meanings are found, the first draft, called an article, is published. This is then proof-read and sent to the General Editor for his first review. Upon finalising, the article of one word is readied and sent to the press. It is once again proof-read by the scholars and the General Editor, before it is finalised as a dictionary entry.” said Sarika Mishra, an Editorial Assistant on the Project.
Publications
Advertisement
While the first volume took three years to be published in 1976, technological intervention and an exclusive software with a font named KoshaSHRI have quickened the process.
“Now, we are able to publish a volume in little over a year. Approximately 4,000 words are incorporated in a volume,” said Onkar Joshi, also an Editorial Assistant of the Project.
In case of any missing information observed in the reference slips, the scholars re-read/scan the 1,464 books, now digitised, effectively making it a double reading of voluminous Mahabharata (18 Parvans), Vedas and alike.
” We can now use the software to easily scan through the books. In the past, this used to be done manually and would be time consuming,” Onkar added.
Since 1976, a total of 6,056 pages of words starting with the first alphabet ‘ अ ‘ have been published in 35 volumes.
Advertisement
” Alphabet ‘ अ ‘ has the maximum words and we have published 35 volumes consisting of words starting from this alphabet. Work on the 36th volume is underway,” said Sanhita Joshi, also an Editorial Assistant of the Project.
Unique and the largest dictionary
Pro Vice-Chancellor Professor Prasad Joshi is the ninth General Editor, and third from the family after his father and uncle, to work in this project.
Advertisement
“This job is a minute-to-minute and day-to-day job,” said Prof Joshi, who has been in-charge of the project since 2017.
Asked if there is any other language in the world that has such a rich and vast vocabulary, he said, “Possibly, the English language dictionary based on historical principles, which took nearly 100 years to be completed, will come close. But the Sanskrit dictionary has wider scope.”
Advertisement
For comparison, the Oxford English dictionary, with 20 volumes and 2,91,500 word entries so far, remains among the most popularly used dictionaries. The Woordenboek Der Nederlandsche Taal (WNT) is another large monolingual dictionary in Dutch. It contains 4.5 lakh words in 17 volumes.
The Encyclopaedic Sanskrit Dictionary, once ready, will be three times larger. The 35 volumes published so far contain about 1.25 lakh vocables (word).
Though there are 46 alphabets in Sanskrit language and several more decades of work lay ahead towards completion of this project, it is estimated that in the end, it will be a dictionary with a total vocabulary of 20 lakh words.
Future
Prof Joshi’s team is the crucial link between the past and the future, and has a big responsibility to keep Sanskrit alive. But there is a real shortfall of Sanskrit linguists.
“Overall, language studies have remained on the backfoot. We need readers for the vast volumes of scriptures and literary works lying unread,” he said.
But young scholars such as Bhav Sharma, Editorial Assistant and Project’s Secretary, are now reaching out to the public aimed to inspire a few.
“We need to showcase to the students the efforts and processes required for dictionary-making. We are planning student-centric activities in the near future so that there is hands-on learning,” said Sharma.
Presently, all the published volumes remain accessible in hard copy format.
The college administration is working aggressively towards making digital copies available within a year.
The Project, KoshaSHRI, under which the website for online access of the Dictionary will be made, also consists of a customised software which is presently under testing and development.
This will speed up the process of Dictionary making in the coming years.
OpenAI has introduced a new automatic speech recognition (ASR) system called Whisper as an open-source software kit on GitHub. Whisper’s AI can transcribe speech in multiple languages and translate them into English, though the GPT-3 developer claims Whisper’s training makes it better at distinguishing voices in loud environments and parsing heavy accents and technical language.
Whisper Writing
Whisper trained its ASR model on 680,000 hours of “multilingual and multitask” data pulled from the web. The idea is that a broad approach to data collection improves Whisper’s ability to understand more speech because of the different accents, environmental noise, and subjects discussed. The AI can understand and transcribe many languages and translate any of them into English. You can see an example in the Korean song translated and transcribed below.
The @OpenAI Whisper speech to text model is multilingual and can even transcribe K-Pop:https://t.co/x8TtZmPR1m pic.twitter.com/PNY3Gs2kjP
— Andrew Mayne (@AndrewMayne) September 21, 2022
While impressive, OpenAI’s research paper suggests the ASR is really only that successful in about 10 languages, a limitation likely stemming from how two-thirds of the training data is in English. And while OpenAI admits Whisper’s accuracy doesn’t always measure up to other models, the “robust” nature of its training puts it ahead in other And though the “robust” training enables Whisper to discern and transcribe speech through background noise and accent variations, it also creates new problems.
“Our studies show that, over many existing ASR systems, the models exhibit improved robustness to accents, background noise, technical language, as well as zero shot translation from multiple languages into English; and that accuracy on speech recognition and translation is near the state-of-the-art level,” OpenAI’s researchers explained on GitHub. “However, because the models are trained in a weakly supervised manner using large-scale noisy data, the predictions may include texts that are not actually spoken in the audio input (i.e. hallucination). We hypothesize that this happens because, given their general knowledge of language, the models combine trying to predict the next word in audio with trying to transcribe the audio itself.”
OpenAI is often in the news for GPT-3 and related products like text-to-image generator DALL-E. Whisper offers a glimpse at how the company’s AI research extends into other arenas. Whisper is open-source, but the value of neural net AI speech recognition for consumer and business purposes is conclusively proven at this point. Whisper could be a starting point for OpenAI to join in, as the researchers already speculated.
“We anticipate that Whisper models’ transcription capabilities may be used for improving accessibility tools. While Whisper models cannot be used for real-time transcription out of the box – their speed and size suggest that others may be able to build applications on top of them that allow for near-real-time speech recognition and translation. The real value of beneficial applications built on top of Whisper models suggests that the disparate performance of these models may have real economic implications.
OpenAI Drops GPT-3 API Prices As Enterprise Adoption Rises
Imagine you are in a foreign country where you don’t speak the language and your small child unexpectedly starts to have a fever seizure. You take them to the hospital, and the doctors use an online translator to let you know that your kid is going to be OK. But “your child is having a seizure” accidentally comes up in your mother tongue is “your child is dead.”
This specific example is a very real possibility, according to a 2014 study published in the British Medical Journal about the limited usefulness of AI-powered machine translation in communications between patients and doctors. (Because it’s a British publication, the actual hypothetical quote was “your child is fitting.” Sometimes we need American-British translation, too.)
Machine translation tools like Google Translate can be super handy, and Big Tech often promotes them as accurate and accessible tools that’ll break down many intra-linguistic barriers in the modern world. But the truth is that things can go awfully wrong. Misplaced trust in these MT tools’ ability is already leading to their misuse by authorities in high-stake situations, according to experts—ordering a coffee in a foreign country or translating lyrics can only do so much harm, but think about emergency situations involving firefighters, police, border patrol, or immigration. And without proper regulation and clear guidelines, it could get worse.
Machine translation systems such as Google Translate, Microsoft Translator, and those embedded in platforms like Skype and Twitter are some of the most challenging tasks in data processing. Training a big model can produce as much CO2 as a trans-Atlantic flight. For the training, an algorithm or a combination of algorithms is fed a specific dataset of translations. The algorithms save words and their relative positions as probabilities that they may occur together, creating a statistical estimate as to what other translations of similar sentences might be. The algorithmic system, therefore, doesn’t interpret the meaning, context, and intention of words, like a human translator would. It takes an educated guess—one that isn’t necessarily accurate.
In South Korea, a young man used a Chinese-to-Korean translation app to tell his female co-worker’s Korean husband they should all hang out together again soon. A mistranslation resulted in him erroneously referring to the woman as a nightlife establishment worker, resulting in a violent fistfight between the two in which the husband was killed, the Korea Herald reported in May. In Israel, a young man captioned a photo of himself leaning on a bulldozer with the Arabic caption “يصبحهم,” or “good morning,” but the social media’s AI translation rendered it as “hurt them” in English or “attack them” in Hebrew. This led the man, a construction worker, to being arrested and questioned by police, according to the Guardian in October 2017. Something similar happened in Denmark, where, the Copenhagen Post Online reported in September 2012, police erroneously confronted a Kurdish man for financing terrorism because of a mistranslated text message. In 2017, a cop in Kansas used Google Translate to ask a Spanish-speaker if they could search their car for drugs. But the translation was inaccurate and the driver did not fully understand what he had agreed to given the lack of accuracy in the translation. The case was thrown out of court, according to state legal documents.
These examples are no surprise. Accuracy of translation can vary widely within a single language—according to language complexity factors such as syntax, sentence length, or the technical domain—as well as between languages and language pairs, depending on how well the models have been developed and trained. A 2019 study showed that, in medical settings, hospital discharge instructions translated with Google Translate into Spanish and Chinese are getting better over the years, with between 81 percent and 92 percent overall accuracy. But the study also found that up to 8 percent of mistranslations actually have potential for significant harm. A pragmatic assessment of Google Translate for emergency department instructions from 2021 showed that the overall meaning was retained for 82.5 percent of 400 translations using Spanish, Armenian, Chinese, Tagalog, Korean, and Farsi. But while translations in Spanish and Tagalog are accurate more than 90 percent of the time, there’s a 45 percent chance that they’ll be wrong when it comes to languages like Armenian. Not all errors in machine translation are of the same severity, but quality evaluations always find some critical accuracy errors, according to this June paper.
The good news is that Big Tech companies are fully aware of this, and their algorithms are constantly improving. Year after year, their BLEU scores—which measure how similar machine-translated text is to a bunch of high quality human translations—get consistently better. Just recently, Microsoft replaced some of its translation systems with a more efficient class of AI model. Software programs are also updated to include more languages, even those often described as “low-resource languages” because they are less common or harder to work with; that includes most non-European languages, even widely used ones like Chinese, Japanese, and Arabic, to small community languages, like Sardinian and Pitkern. For example, Google has been building a practical machine translation system for more than 1,000 languages. Meta has just released the No Language Left Behind project, which attempts to deploy high-quality translations directly between 200 languages, including languages like Asturian, Luganda, and Urdu, accompanied by data about how improved the translations were overall.
However, the errors that lead to consequential mistakes—like the construction worker experienced—tend to be random, subjective, and different for each platform and each language. So cataloging them is only superfluously helpful in figuring out how to improve MT, says Félix Do Carmo, a senior lecturer at the Centre for Translation Studies at the University of Surrey. What we need to talk about instead, he says, is “How are these tools integrated into society?” Most critically, we have to be realistic about what MT can and cannot do for people right now. This involves understanding the role machine translation can have in everyday life, when and where it can be used, and how it is perceived by the people using it. “We have seen discussions about errors in every generation of machine translation. There is always this expectation that it will get better,” says Do Carmo. “We have to find human-scale solutions for human problems.”
And that means understanding the role human translators still need to play. Even as medications have gotten massively better over the decades, there still is a need for a doctor to prescribe them. Similarly, in many translation use cases, there is no need to totally cut out the human mediator, says Sabine Braun, director of the Centre for Translation Studies at the University of Surrey. One way to take advantage of increasingly sophisticated technology while guarding against errors is something called machine translation followed by post-editing, or MT+PE, in which a human reviews and refines the translation.
One of the oldest examples of a company using MT+PE successfully is detailed in this 2012 study about Autodesk, a software company that provides imaging services for architects and engineers, which used post-editing for machine translation to translate the user interface into 12 languages. Other similar solutions have been reported by a branch of the consulting company EY, for example, and the Swiss bank MigrosBank, which found that post-editing boosted translation productivity by up to 60 percent, according to Slator. Already, some machine translation companies have stopped selling their technologies for direct use of clients and now always require some sort of post-editing translation, Do Carmo says. For example, Unbabel and Kantan are platform plugins that businesses add into their customer support and marketing workflows to reach clients all over the world. When they detect poor quality in the translated texts, the texts are automatically routed to professional editors. Although these systems aren’t perfect, learning from these could be a start.
Ultimately, Braun and Do Carmo think that it’s necessary to develop holistic frameworks that go far beyond the metrics used at the moment to assess or evaluate quality of translation, like BLEU. They would like to see the field working on an evaluation system which encompasses the “why” behind the use of translation, too. One approach might be an independent, international regulatory body to oversee the use and development of MT into the real world—with plenty of social scientists on board. Already, there are many standards in the translation industry as well as technological standardization bodies, like the W3 organization—so experts believe it can be done, as long as there is some more organization in the industry.
Governments and private companies alike also need clear policies about exactly when officials should and should not use machine translation tools, either free consumer ones or others. Neil Coulson is the founder of CommSOFT, a communication and language software technology company trying to make machine translation safer. “Police forces, border-control agencies, and many other official organizations aren’t being told that machine translation isn’t real translation and so they give these consumer gadgets a go,” he says. In March 2020, his organization sent out a Freedom of Information request to 68 different large U.K. public-sector organizations asking for their policies on the use of consumer gadget translation technologies. The result: None of these organizations had any policy for their use of MT, and they do not monitor any of their organizational or staff’s ad-hoc use of MT. This can lead to an unregulated, free-for-all landscape in which anyone can publish a translation app and claim that it works, says Coulson. “It’s a ‘let a thousand flowers bloom’ approach … but eventually someone eats a flower that turns out to be poisonous and dies,” says Coulson.
Education about the pros and cons of MT, of course, is paramount—among researchers, companies, and organizations who want to actually start using the tool, but most importantly, among everyday users. That’s why Lynne Bowker, a professor of translation and interpretation at the University of Ottawa, started the Machine Translation Literacy project. Their goal is to spread awareness of how MT systems process information and teach researchers and scholars how to actually use MT more effectively. Including information about machine translation as part of the broader digital literacy and information literacy training given to school kids would also be welcome. “Being machine translation literate means understanding the essentials of how this technology works in order to be able to evaluate its strengths and weaknesses for a particular task or use,” says Bowker. Language, in a social context, is communication. “One of the real challenges we are facing is how to reach the wider public with this message,” says Bowker.
Being able to differentiate between low-stakes tasks and high-stakes tasks remains one of the key points, Bowker says. Thankfully, in the meantime, most mistranslations still just lead up to a laugh: According to a 2016 study in International Journal of Communication, there’s a Chinese restaurant called Translate Server Error. Tthe MT system mistranslated the original language, but the restaurant owners didn’t know English well enough to realize something was off.
Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.
Terms including “plant-based” and “oat milk” have officially been added to the Merriam-Webster dictionary.
Plant-based, which is thought to have been first used in language in 1960, is given two definitions in the dictionary.
The first is “made or derived from plants,” and the dictionary cites “plant-based burger” as an example. The second definition is “consisting primarily or entirely of food (such as vegetables, fruits, nuts, oils, and beans) derived from plants.”
Merriam-Webster defines oat milk as “a liquid made from ground oats and water that is usually fortified (as with calcium and vitamins) and used as a milk substitute.” The first known use of the term oat milk was in the 1980s.
The two terms fall under the food category, and have been added alongside the popular autumn flavor “pumpkin spice.”
In addition to this, Merriam-Webster has also added “greenwashing.” This denotes the process of attempting to make something appear more environmentally friendly than it actually is.
'Greenwashing' is now entered, defined as "the act or practice of making a product, policy, activity, etc. appear to be more environmentally friendly or less environmentally damaging than it really is."https://t.co/KHi9OCBTB5
— Merriam-Webster (@MerriamWebster) September 7, 2022
The history of Merriam-Webster
Merriam-Webster has been publishing dictionaries since 1847, and first launched its online version in 1996.
The dictionary is regularly updated with new words and phrases, and in recent years additions have included “selfie,” “cancel culture,” and “hygge.”
This year’s 370 new words and terms also include “booster dose,” “sus,” and “lewk.”
“The dictionary chronicles how the language grows and changes, which means new words and definitions must continually be added,” reads a statement from Merriam-Webster. “When many people use a word in the same way, over a long enough period of time, that word becomes eligible for inclusion.”
The rise of the plant-based movement
The new additions were undoubtedly prompted by the rise of plant-based eating. This year saw a huge 629,000 people sign up for Veganuary, up from 580,000 the previous year.
According to research from 2021, there are approximately 79 million vegans in the world.
Oat milk has also seen a staggering increase in popularity, with sales doubling from 2019 to 2020. The oat milk industry alone has been forecast to be worth $6.45 billion by 2028.
The drink is hugely popular among vegans and non-vegans alike.
A study published earlier this year even found that almost half of Generation Z (those born between 1997 and 2012) felt shame when ordering dairy milk. The study also found that more than half intended to give it up over the next year.
:quality(70)/cloudfront-us-east-1.images.arcpublishing.com/adn/EGTPS37FUFHCTNI75U527KUVEQ.jpg)
OpenAI has introduced a new automatic speech recognition (ASR) system called Whisper as an open-source software kit on GitHub. Whisper’s AI can transcribe speech in multiple languages and translate them into English, though the GPT-3 developer claims Whisper’s training makes it better at distinguishing voices in loud environments and parsing heavy accents and technical language.
