AI has made another assault on the English language. After taking the title of the Collins Dictionary word of the year, artificial intelligence this week assailed the Cambridge version.
This time, the impact is more subtle — but deeper. The first example emerged from Cambridge’s word of the year for 2023: “hallucinate.”
It’s an old word, but the award is due to a new meaning. In the latest versions of the Cambridge Dictionary, “hallucinate” has an extra definition:
“When an artificial intelligence (= a computer system that has some of the qualities that the human brain has, such as the ability to produce language in a way that seems human) hallucinates, it produces false information.”
The <3 of EU tech
The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!
To clarify the concept, the entry also includes two examples:
LLMs are notorious for hallucinating — generating completely false answers, often supported by fictitious citations.
The latest version of the chatbot is greatly improved but it will still hallucinate facts.
Hallucinating words
At TNW Towers, we welcome Cambridge’s intervention. Tech experts may argue that they’re reducing hallucinations, but the problem is far from solved — and the results can be alarming.
ChatGPT, for instance, can spoutdangerous medical advice. Security analysts fear the bot’s hallucinations could also drive malicious code towards software developers.
There are also risks for the Cambridge Dictionary.
“Managing the tendency of generative AI tools to hallucinate will be key to ensuring our users can continue to trust us,” said Wendalyn Nichols, the book’s publishing manager.
“The emergence of a new meaning of hallucinate is a great case in point. It’s human experts tracking and capturing changes in the language that make the Cambridge Dictionary a trustworthy source of information about new words and senses — ones the public-facing AI tools won’t have learned yet.”
AI’s second linguistic influence
The new meaning of hallucinate isn’t the only mark of AI on the Cambridge lexicon.
During 2023, lexicographers have added various definitions related to artificial intelligence, including large language model (or LLM), generative AI (or GenAI), and GPT.
At the tech’s current pace of development, the impact on 2024’s dictionaries could be even deeper.
With the generative AI explosion still booming, perhaps artificial intelligence will simply invent the next word of the year.
Iranian Christians attend the New Year mass at the Saint Targmanchats Armenian Church in Tehran on Jan. 1, 2023. | Atta Kenare/AFP via Getty Images
As the underground Church continues to grow in Muslim-majority Iran, Bible translators are putting their lives on the line to bring the Gospel into the local dialects so that their friends and neighbors can have access to the written word of God for the first time.
Through the work of the translation agency unfoldingWord, Christians in Iran and across the globe have been able to translate the Gospel themselves into more native languages.
The Christian Post interviewed a representative from unfoldingWord, Evan Thompson, who preferred to use a fake name or pseudonym for his safety.
"There are 1.45 billion people in the world who speak about 5,500 languages that do not have the whole Bible in their heart languages. ... The Church has expanded exponentially in the last 20 years. And what these folks have learned is that you can lead someone to Christ, but if they don't have a church, they don't survive on their own," Thompson said.
"You can start a church, but if that church doesn't have the Bible in its heart language, it will typically only last one generation. Iran, for example, has churches operating underground. And there are thousands of underground churches in many other parts of the world," he added.
UnfoldingWord, a nonprofit organization that has been around for roughly seven years, "works with Church leaders around the world who are seeking to establish their churches in sound doctrine, but lack access to Bible translations in the languages their people speak."
A day in the life of Iranian Bible translators
The Christian Post heard from two Iranian women risking their lives to help translate unfoldingWord's Open Bible story resources from Farsi into other Iranian dialects for evangelizing.
Both women Bible translators have chosen not to give their real names for this article to protect their identities and maintain their safety.
The first woman, using the name Miriam, said she gave her heart to Christ after coming to the realization that she is "God's child and daughter."
Miriam is part of a people group in Iran that is made up of millions of natives. She says she is often treated as a second-class citizen because of how those from other people groups view her status in her people group.
Miriam's life could be in danger if the Iranian government finds out that she follows Jesus in the Islamic Republic, which Open Doors ranks as the eighth-most hostile country for Christians.
"God is my Father. I feel deeply honored to be part of this work of bringing God's Word to my people," Miriam said.
Despite having children and knowing that her life is at risk for believing in Jesus, Miriam said she will not stop working to translate the Gospel into her heart language.
"I cannot even imagine leaving this work unfinished. I must complete this work and see the result. I want to see my beloved ones experience salvation in Christ. This is my dream; that my people can talk about God and speak His name freely without any hesitation; without any fear they can talk about God," she said.
Miriam was introduced to Christianity through a friend in college who gave her a Farsi New Testament. She had to read the Bible alone and in secret, an act that left her without much of a clear understanding of the Christian faith.
After college, Miriam married into a strict Muslim family. But, no matter how hard she tried to adapt to the strict religious practices of Islam, she could not find God as a Muslim.
Miriam said she didn't give her life entirely to Jesus until after she heard about Transform, an online class offered in Iran that covered the basic teachings of Christianity.
She watched the classes secretly through various digital platforms. And during one of the class sessions, she gave her life to Christ.
Following her conversion, Miriam's husband caught her one day watching the Transform Iran pastor on television.
Miriam could no longer hold back the truth about her faith from her husband.
"By the grace of God, he did not get angry. He said, 'I know you are a serious-minded woman, and if this is important to you, it's OK,'" Miriam recalled.
Miriam's husband began watching the class with her, and several months later, he also gave his life to Christ.
Before her husband's conversion, the Transform Iran pastor asked her if she would get involved in Bible translation because of her expertise in her heart language.
Miriam accepted the offer even if that meant risking her life to help translate the Bible into more Iranian tongues.
"We are not allowed to study our heart languages in Iranian public schools. This is a limitation for our people. I have this language specialty and experience, this expertise so that I can help my own people. People like my mother can read this book," Miriam said.
"I have a Bible in Farsi, and I can read it. But I cannot understand the more complicated concepts in it because Farsi is not my heart language. I couldn't establish a relationship with the Bible in Farsi. I'm very fluent in Farsi. I studied hard and had great teachers. Still, I cannot establish a relationship with the Bible in Farsi," she continued.
"How about other people who don't have my educational advantages? My family and friends? Having the Gospel in my heart language makes it much easier to talk to my family about Jesus. They can understand and accept Him easily."
'Jesus fed me'
Another Iranian Bible translator who is using the pseudonym Stella accepted Jesus into her heart after her husband died of cancer.
Following her husband's death, Stella was left alone to care for her young son. During that time, she relied on the peace of God as her only hope.
"God has helped me. The Name of Jesus Christ was in my life. I didn't need anybody. Jesus fed me, put clothes on me and gave me peace," she said.
Stella learned more about God through a Bible translated into Farsi. At first, she thought Christianity was a religion. But, now she understands that Christianity is a relationship.
"When I was a new believer, I was thinking that, 'OK, I'm going to just switch religions.' But, when I got to know the Holy Spirit, I understood that this is a relationship, not religion," Stella said.
Stella is currently working on a Bible translation in her heart language. Her sister-in-law became a Christian thanks to her work translating the Bible. Stella worked for five years alongside her family as they helped her review the translation of the Bible, and now she works as part of a larger Bible translating group.
"I love my mother language. I'm telling the poetry; I write the context. I write the sentence. I record it. … I know all of this is God's work for us. God wants us to do this. … I am thinking about my mom, my father, my childhood. And everyone that doesn't have it right now. I really want to bring God to my town and my people," she said.
A dire need for Bible translations
Before unfoldingWord was launched, traditional Bible translation agencies across the globe have done "marvelous work" and continue to do so, Thompson noted.
However, he said the number of Western Bible translators that Bible translation agencies can send overseas is dropping, and the demand for Bible translation is increasing rapidly.
"The group that founded unfoldingWord developed a way to address this problem. We call it church-centric Bible translation. ... It's Bible translation incorporated into the life of the Church as part of its discipleship," Thompson noted.
"Most of those unreached people have neighbors who know Christ, and they're taking the Gospel to them. And what unfoldingWord does is we equip the Church in every people group with a goal of translating the Bible in every language."
To help local churches translate the Bible, unfoldingWord provides people groups with open-source software and open-licensed biblical content that's breaking the copyright barriers to source texts.
The organization also has comprehensive translation guides to answer difficult Bible translation questions. The organization offers essential doctrinal education to protect the theological integrity of the translations.
"unfoldingWord provides training for indigenous Bible translation teams over Zoom and sometimes in neutral locations," Thompson said.
The training locations are kept secret to protect participants in certain countries where practicing Christianity is not accepted.
"Our training allows for indigenous Bible translation teams to be able to use best practices when they are translating the Bible for themselves. One of the ways that I like to say it is: 'We don't make Bible translations. We help develop Bible translators.' Because that's what the Church really needs all over the world."
Issues bringing translations to Iran
Thompson said unfoldingWord has aided Iranian natives who have translated the Bible into dozens of native languages.
However, he said there are some unfortunate limitations to the organization's ability to aid translators in Iran. He cited strict government policies prohibiting Iranians from studying their heart languages or their native tongues in public schools.
"All of these oppressive countries, like Sudan, like Iran, and some others we could name, are trying to Islamasize their whole population. And one of the ways that they do that is by forcing them to speak this one national language, and it's causing their heart languages to die out," Thompson said.
In Iran, the government recognizes Farsi as the national language. However, many natives speak other Iranian dialects more fluently.
"It's very much like anybody that comes to America from someplace else, and we put them in American schools to learn English. Unless their family makes it a point to keep their native languages ... alive in their families, by the second or third generation, the kids can't speak their native language anymore," Thompson said.
The few responsible for the many
Miriam believes God has blessed her with a huge responsibility to help translate the Bible into more Iranian languages.
"This is not just a scientific book. This is God's Word. I felt some tension. I was scared that I would not do the work well enough, but I'm very happy to make it available so that my people can establish a relationship with the Bible. That's why I got involved with this Bible translation project," she said.
When asked if she can imagine the day when the Bible is completed in even more Iranian languages, Miriam said it will take many years, and she doesn't know if she'll live long enough to see the project's conclusion.
"I want my children to experience the result of my work so that they can know Jesus through their heart language. I never thought this project would grow like this. But I've learned that it's not just about me," Miriam said.
"I need this team that has been pulled together. For safety, we have many people working on this project secretly. There may be several Christians in one Iranian family, but they cannot share their faith with each other openly," she continued.
"We need more people who can speak my heart language to continue working on this project. … I want to ask for prayer … to recruit people into the project who know our language well."
The days of having a dictionary on your bookshelf are numbered. But that’s OK, because everyone already walks around with a dictionary – not the one on your phone, but the one in your head.
Just like a physical dictionary, your mental dictionary contains information about words. This includes the letters, sounds and meaning, or semantics, of words, as well as information about parts of speech and how you can fit words together to form grammatical sentences. Your mental dictionary is also like a thesaurus. It can help you connect words and see how they might be similar in meaning, sound or spelling.
As a researcher who studies word retrieval, or how you quickly and accurately pull words out of your memory to communicate, I’m intrigued by how words are organized in our mental dictionaries. Everyone’s mental dictionary is a little bit different. And I’m even more intrigued by how we can restore the content of our mental dictionaries or improve our use of them, particularly for those who have language disorders.
Language is part of what makes humans special, and I believe everyone deserves the chance to use their words with others.
Your mental dictionary
While a physical dictionary is helpful for shared knowledge, your personal mental dictionary is customized based on your individual experiences. What words are in my mental dictionary might overlap with the mental dictionary of someone else who also speaks the same language, but there will also be a lot of differences between the content of our dictionaries.
You add words to your mental dictionary through your educational, occupational, cultural and other life experiences. This customization also means that the size of mental dictionaries is a little bit different from person to person and varies by age. Researchers found that the average 20-year-old American English speaker knows about 42,000 unique words, and this number grows to about 48,000 by age 60. Some people will have even larger vocabularies.
By now, you might be envisioning your mental dictionary as a book with pages of words in alphabetical order you can flip through as needed. While this visual analogy is helpful, there is a lot of debate about how mental dictionaries are organized. Many scholars agree that it’s probably not like an alphabetized book.
One widely rejected theory, the grandmother cell theory, suggests that each concept is encoded by a single neuron. This implies that you would have a neuron for every word that you know, including “grandmother.”
While not accepted as accurate, the aspect of the grandmother cell theory suggesting that certain parts of the brain are more important for some types of information than others is likely true. For example, the left temporal lobe on the side of your brain has many regions that are important for language processing, including word retrieval and production. Rather than a single neuron responsible for processing a concept, a model called parallel distributed processing proposes that large networks of neurons across the brain work together to bring about word knowledge when they fire together.
For example, when I say the word “dog,” there are lots of different aspects of the word that your brain is retrieving, even if unconsciously. You might be thinking about what a dog smells like after being out in the rain, what a dog sounds like when it barks, or what a dog feels like when you pet it. You might be thinking about a specific dog you grew up with, or you might have a variety of emotions about dogs based on your past experiences with them. All of these different features of “dog” are processed in slightly different parts of your brain.
Using your mental dictionary
One reason why your mental dictionary can’t be like a physical dictionary is that it is dynamic and quickly accessed.
Your brain’s ability to retrieve a word is very fast. In one study, researchers mapped the time course of word retrieval among 24 college students by recording their brain activity while they named pictures. They found evidence that participants selected words within 200 milliseconds of seeing the image. After word selection, their brain continued to process information about that word, like what sounds are needed to say that chosen word and ignoring related words. This is why you can retrieve words with such speed in real-time conversations, often so quickly that you give little conscious attention to that process.
Until … you have a breakdown in word retrieval. One common failure in word retrieval is called the tip-of-the-tongue phenomenon. It’s the feeling when you know what word you want to use but are unable to find it in that moment. You might even know specific details about the word you want, like other words with similar meaning or maybe the first letter or sound of that word. With enough time, the word you wanted might pop into your mind.
These tip-of-the-tongue experiences are a normal part of human language experience across the life span, and they increase as you grow older. One proposed reason for this increase is that they’re due to an age-related disruption in the ability to turn on the right sounds needed to say the selected word.
For some people, though, tip-of-the-tongue experiences and other speech errors can be quite impairing. This is commonly seen in aphasia, a language disorder that often occurs after injury to the language centers of the brain, such as stroke, or neurodegeneration, such as dementia. People with aphasia often have difficulty with word retrieval.
Fortunately, there are treatments available that can help someone improve their word retrieval abilities. For example, semantic feature analysis focuses on strengthening the semantic relationships between words. There are also treatments like phonomotor treatment that focus on strengthening the selection and production of speech sounds needed for word production. There are even apps that remotely provide word retrieval therapy on phones or computers.
The next time you have a conversation with someone, take a moment to reflect on why you chose the specific words you did. Remember that the words you use and the mental dictionary you have are part of what make you and your voice unique.
This article was originally published on The Conversation. Read the original article.
Since Scrabble adopted an official lexicon in 1978, one thing has been constant: People have never stopped arguing about what is or isn’t a word.
Players have defended the game by noting that its letter strings—from AA (a kind of Hawaiian lava) to ZZZ (an interjection for sleep)—could be found in a bunch of standard North American dictionaries, books that have been used through the years to compile and revise Scrabble’s tournament word list. But after an update last month introduced dozens of suspect words, riling up the community of competitive players, that’s becoming harder to do.
The linguistic tumult began in September, when NASPA Games, the organization that maintains the word list used in club and tournament Scrabble, published a draft of its update. The NASPA list includes all of the words in the Official Scrabble Players Dictionary, the go-to source for living-room and app players in North America, plus a lot more. Although the seventh edition of the OSPD, which I wrote about last year, ushered in more than 500 newbies, NASPA added more than 4,700.
The big initial story was a flip-flop: NASPA said it was reinstating more than 100 of around 250 slurs removed in 2020 amid a fractious debate prompted by the murder of George Floyd and the Black Lives Matter protests. The slurs were deleted according to two standards the group created: strict (a word is judged by all dictionary sources to be at least “often offensive”) and less-strict (a word is at most only “sometimes offensive” and/or has an alternative, inoffensive meaning). But the less-strict standard now “seemed too far removed from our original intention,” NASPA chief executive John Chew told me in an email. So the group re-added words such as SPAZ, GOY, REDNECK, GREYBEARD, and TWAT. “We don’t like removing words from play,” Chew said.
Some Scrabble experts, though, were less concerned with definitional hairsplitting than the optics of retreating from a moral stand. “The fact that this issue is being cast back into the limelight, and that players now have to relearn a set of words they are now programmed to know as unacceptable slurs, is completely counterproductive to the initial mission of expurgation—namely, getting rid of these slurs so people wouldn’t have to encounter or think about them,” Mack Meller, the No. 3–ranked player in North America, said in an email. Top-ranked Josh Sokol, who in July won the 2023 North American championship run by NASPA, told me, “Nobody was asking to reinstate half of the slurs. Nobody.”
But the slurs were just a warmup act. When players, including me, started combing the list of additions of words up to 15 letters long—the OSPD stops at eight—they found a bunch of head-scratching stuff, mostly involving the inflected forms of words, with the plural endings -s and -es and the comparative and superlative endings -er, -est, -ier, and -iest.
Inflections can be tricky. Dictionaries have rarely listed every inflected form of root words—in the print age, for reasons of space and expense; now, online, because of convention. So Chew and his colleagues on NASPA’s dictionary committee did what Scrabble players have done since the first OSPD was published by Merriam-Webster Inc. 45 years ago: They tried to apply rules enumerated by dictionaries to guide decisions on the validity of a word.
That’s a sensible approach, and NASPA spent hundreds of hours working on the update; a report to players cites numerous dictionary sources and drops authoritative-sounding terms like cutback plural and suffixal identity. Chew, thanks in part to his Scrabble work, was recently named editor of a planned dictionary of Canadian English. But he’s not a professional lexicographer, and NASPA didn’t consult any lexicographers to vet its list. As a result, pro lexicographers told me, the group committed errors in assumption and interpretation about dictionary practices, with some lexically comical results.
Take ROUXES, which NASPA added as a plural of the French-derived cooking agent roux. While Merriam-Webster and four other Scrabble source dictionaries explicitly state that the plural of roux is roux, the fourth edition of Webster’s New World College Dictionary, published in 1999, does not specify a plural.* Chew told me that dictionary’s front-matter rule regarding plurals of words ending in -s, -x, -z, and -sh is “extremely systematic” and calls for an -es ending when no plural is specified, which “unambiguously supports ROUXES.”
Steve Kleinedler, the managing editor of the fifth edition of Webster’s New World (2014), said NASPA’s interpretation is incorrect.* That particular rule about plurals, he explained, is intended for words that end in the sibilant sounds of the letters—like box or tax—not for French-derived words in which the x is silent. “They’re reading way too much in the front matter of a collegiate dictionary that is simplifying broadly so that it’s not bogged down in describing a handful of edge cases,” Kleinedler emailed. “I don’t think this (mis)reading of Webster’s New World should be taken as a basis for adding words.”
Or consider FECESES. According to NASPA, it was added because the second edition of the Canadian Oxford Dictionary (2005) doesn’t list a plural for feces, even though numerous other dictionaries specify, and common sense confirms, that feces is a plural noun. In his email, Chew said that when there is doubt about implied inflections, “we look for credible published citations to support them.” In a database for members, NASPA lists three citations for FECESES: a story in Fox News (the only Google hit I found from mainstream media), a passage in a textbook about Japanese marine life, and a 1934 article in a magazine about fox and fur ranching.
Lexicographers use examples of words in the wild to help determine whether and when a term should be entered into a dictionary. The general principle: widespread use in professionally edited publications over a long period of time. “As lexicographers know too well, you can find just about any form out there if you look long enough, but that doesn’t necessarily make the form dictionary-worthy,” Ben Zimmer, the language columnist for the Wall Street Journal and a former editor of American dictionaries at Oxford University Press, told me. A search of the NOW Corpus, a database of online newspapers and magazines that’s updated daily and includes more than 18 billion words, yields zero hits for FECESES. It’s just not a valid word.
Errors have always sneaked into both standard and Scrabble lexicons—around 30 disputed inflections are being removed in this update; Godspeed, NUMBNUTSES.
Ditto for MIREPOIXS, an inferred plural of the borrowed French word meaning a mixture of diced vegetables used in soups and stews; NASPA endorses its addition with a link to an online recipe for “boho carrot ginger seasoning cubes.” And DEBRISES, which is supported by a citation from the American-Eurasian Journal of Agricultural & Environmental Science. And HORSEFEATHERSES, which will be playable because an edition of Webster’s New World doesn’t specify that horsefeathers is a plural noun meaning, um, nonsense.
Other weird inflections—BRUTALER, JUVENILIAS, SUBSPECIESES—are justified because they appear in a drop-down when typed into the search bar on Merriam-Webster’s homepage. But, lexicographers told me, that’s not a reliable guide. “It’s common practice in all sorts of web search architectures to include false positives in the hopes that you’ll funnel users towards valid data,” said Kory Stamper, Dictionary.com’s senior editor of lexicography.
Merriam-Webster senior editor Emily Brewster told me there’s no easy way to determine from the dictionary’s website which inflections editors consider worthy. “Some forms are not included in the drop-down because the entry has not yet been updated,” she said. “Others are included so that users will find an entry, even if we don’t have sufficient evidence to consider the word established and therefore eligible for play” in Scrabble.
Merriam-Webster editor-at-large Peter Sokolowski, a colleague of Brewster’s, said the drop-down is reliable for recent entries, in which inflections are included or omitted as an editorial choice. I asked him about broey, an adjective meaning like a bro, “a young male who is part of a group of similar male friends stereotypically characterized as hearty, athletic, self-confident, party-loving, etc.” Merriam-Webster added broey online in 2018 and to the OSPD last year. NASPA inferred the inflections BROIER and BROIEST. But those words do not appear when typed into the Merriam search bar—because, Sokolowski said, editors specifically decided they’re not established words.
Finally, here’s a “new” “word” that’s prompted a lot of eye-rolling among Scrabblers: GRATINEEED—yes, with three consecutive E’s. NASPA notes that there is an “explicit” entry for the word in the Canadian Oxford, which is true. But a glance at the printed page might signal a problem. Gratinée is spelled with an accent over the first e. The triple-E past tense is entered as gratineéed—with the accent over the second e, suggesting that the first e might be extraneous. In other words, Zimmer told me, it’s probably a typo.
Chew, the NASPA executive, defended the group’s work. Adding words to a list for a board game, he said, is different from entering words in a traditional dictionary. “We have a history of permissivity in our game … in recognizing inflections in the absence of evidence to the contrary,” he said. Because the Scrabble list has grabbed words from differently edited dictionaries over multiple decades, he said, it can be challenging to “create a sense of consistency,” especially “when the needs of our players to know exactly how words are inflected runs up against a regular lexicographer’s need to edit to space and budget.”
“If what regular lexicographers did had direct application to the editing of a tournament word list, then we wouldn’t need to edit our own!” Chew wrote.
But as complicated or inconsistent as grammar and dictionary rules might be, professional lexicographers could help decipher guidelines and evaluate individual words. Until a few years ago, a Merriam-Webster editor reviewed the lists that Scrabble players compiled by scouring dictionary sources. Errors have always sneaked into both standard and Scrabble lexicons—around 30 disputed inflections are being removed in this NASPA update; Godspeed, NUMBNUTSES and RHINOCEROI—but the cooperation gave the game a stamp of authority that amateur word sleuthing lacks.
Another benefit of consulting the pros: combating the impression that competitive Scrabble just makes up words. “I gotta admit that I’m not looking forward to explaining … MIREPOIXS to a new player who already might be predisposed to thinking we play with gibberish,” one top expert, Kenji Matsumoto, wrote on Facebook.
The word fight is a distraction at a time when players are making traction in boosting Scrabble’s popularity online. YouTube content creator Kevin Peterson, who goes by T1J, this month posted an engaging 18-minute video, “I Tried to Master Scrabble in 30 Days.” Josh Sokol, the current champ, describes himself as a “Scrabble influencer” and is building a following for his livestreams and videos. No. 3 Mack Meller also makes videos; he’s currently playing a 100-game match against a Scrabble bot.
The most influential Scrabbler online is 2017 North American champion Will Anderson. His YouTube channel has more than 26,000 subscribers, and one of his videos, “The Greatest Scrabble Player Ever Is Underrated,” has racked up more than 1.2 million views. Anderson, who works for Scopely Inc., the creator of the licensed app Scrabble GO, would like the North American game to abandon NASPA’s proprietary list and adopt the OSPD. He said the Merriam-Webster lexicon is used for around 57 percent of turns played in English on Scrabble GO. A more expansive international-English list published by the British dictionary maker Collins accounts for 42.5 percent. NASPA’s list? Just 0.54 percent.
“Why are we going rogue?” Anderson said of competitive Scrabble, which currently has about 1,000 active tournament players in the United States and Canada. “We are deviating from a standard in widespread use across the country, and even the world, by players we are trying to convert into our ranks.”
Part of the problem is that the Scrabble ranks are fractured. The game’s owner in North America, Hasbro Inc., terminated a license with NASPA in 2021 and has no involvement in tournament play. (It does sponsor an annual youth championship, which, full disclosure, I’ve helped organize the past two years.) NASPA also doesn’t currently have a formal relationship with Merriam-Webster, which is licensed by Hasbro to publish the OSPD. There’s a splinter players organization that uses the pre-2020 NASPA word list with all the slurs. And some events use the Collins list.
NASPA expects its updated lexicon to take effect in tournaments in early 2024. In the meantime, Scrabble players will continue to question—and snark about—spurious words. “I predict that in 10,000 years archaeologists will unearth a copy” of NASPA’s list, said another high-ranking player, Carl Johnson, “and cite HORSEFEATHERSES as a critical moment in the collapse of Western civilization.”
Correction, Nov. 15, 2023: This article originally misstated that the fourth and fifth editions of Webster’s New World College Dictionary were published in 2014 and 2020, respectively.
Carmen Rodriguez, seated right, and other parents meet with state Sen. Anthony Portantino in the State Capitol.
Credit: Courtesy of Innovate Public Schools
Do you count on EdSource? If so, please make your donation today.
When Los Angeles mother Tania Rivera signed a crucial document for her son Luis’ special education program in 2022, she was hoping he would be able to return to in-person classes after two years of distance learning.
But the individualized education program, or IEP, required for all children who need special education, was available only in English. Rivera’s first language is Spanish.
Later she was told Luis, who has autism, would have to continue with online learning because the document did not specify that he needed in-person classes. In addition, she says, the document removed his occupational therapy for handwriting because a language interpreter erroneously said she objected.
“It is a big disadvantage that we have, because I have some English, but it is very basic,” Rivera said in Spanish. “If we’re talking about educational terms or legal terms, the meaning can be lost with just one word” mistranslated.
Monthslong waits and faulty or incomplete translations of special education documents are widespread across California for parents who speak languages other than English, according to special education advocates. They say these problems violate parents’ rights to participate in their children’s education plans under the Individuals with Disabilities Education Act, the federal law that regulates special education.
A proposal in the state Legislature, Senate Bill 445, aims to solve some of these problems, but its fate remains uncertain because of concerns over potential cost.
“I’ve never seen a timely translation and I’ve never seen all documents being fully translated,” said Lisa Mosko Barros, founder of SpEducational, an organization that works to educate parents to be advocates for their children with special needs and improve their access to high-quality education. Mosko Barros has worked with dozens of families in Southern California, including Rivera, and trained hundreds of others on navigating the IEP process.
She said she has heard the same complaints over and over.
“I literally spoke to one parent this morning in the Inland Empire who a couple of years ago signed an IEP and didn’t realize she was signing consent to eliminating speech services for her child who is non-verbal with autism,” Mosko Barros said. “It really can make or break a child’s access to a free and appropriate public education.”
Rivera’s son Luis, now in eighth grade, remained in online classes since fifth grade until this fall and regressed as a result, his mother said.
In total, he lost three years of in-person classes, first in 2020-21 when all students had distance learning, again in 2021-22 because he has chronic asthma and his pediatrician recommended he stay home since vaccinations against Covid-19 were not yet available for children. Then, in 2022, the translation problems kept him out of in-person schooling for another year.
“He has had academic setbacks, and socially, he regressed a lot because it was three years without interaction,” Rivera said.
When asked how long the district takes to translate special education assessments and IEP documents, the Los Angeles Unified School District communications team wrote that “the District works to parallel the IEP timeline for consistency and return the translated document within the same 30-day timeframe.” They declined to comment on Rivera’s case.
Rivera and almost 200 other people attended an online meeting in September with state Sen. Anthony Portantino, D-Burbank, at which parents shared how long wait times and poor-quality translations have hurt their children with special needs. They expressed their support for Portantino’s bill, which would require IEPs to be translated into a parent or guardian’s native language by a “qualified translator” within 30 calendar days of an IEP meeting or a later request.
Current federal and state laws require that school districts “take any action necessary” to ensure parents understand IEP meetings, and state law requires they translate a student’s IEP at a parent’s request, but no time frame is specified.
“I believe strongly that parents can best advocate for their children when they have the knowledge to do so. Not being able to read an IEP because of language barriers is unacceptable,” Portantino said. “We must find a way to translate IEPs more quickly.”
Portantino said the issue is personal for him because he struggled with dyslexia and ADHD as a student and received limited help from the schools he attended.
“I largely depended on developing my own learning methods, which included lots of repetition and good listening skills,” Portantino said. But he wants to make sure other children can get the help they need.
The bill passed the Senate, the Assembly Education Committee and the Assembly Appropriations Committee with no opposition. But an analysis by the Assembly Appropriations Committee found that the bill could cost the California Department of Education $409,000 annually and could cost school districts between $6 million and $16 million, which might also have to be reimbursed by the state. Believing there was a risk the bill could be vetoed this year because of those costs, Portantino said he chose to make it a “two-year bill,” giving it more time to be discussed in the Legislature and with Gov. Gavin Newsom.
San Francisco Unified School District passed a policy in 2022 to ensure “every effort shall be made” to translate special education documents before meetings so that parents have time to read and understand them. It also requires meeting times to be extended to allow for interpretation.
Carmen Rodríguez is one of dozens of parents who pushed for that policy. Rodríguez has two children with disabilities. Before the San Francisco Unified policy passed, she said, she waited eight months for a written translation of the first assessment of her older son, who has anxiety and a learning disability, and a year for the IEP for her younger child, who has dyslexia.
“If it’s not in my language, how am I going to understand the document? How do I know that it really says here what my child needs?” Rodríguez said in Spanish.
In addition, she said IEP meetings were often cut short because the district limited them to one hour, with no extra time allowed for interpretation.
Belén Pulido Martínez, senior community organizer for Innovate Public Schools, an organization that worked to get the San Francisco policy passed, said the policy empowers parents.
“Now in San Francisco, the district is training their special ed teachers on the policy, and we’re super happy about that because it’s not just a piece of paper that’s going to die in an office. It’s being implemented,” Martínez said.
Matt Alexander, the San Francisco Unified Board of Education commissioner who worked with parents to write the policy, said school districts have to prioritize translation and interpretation if they want parents to be engaged.
“In our district, over half of our families don’t speak English at home. So if we care about communicating with our families, we have to provide interpretation,” Alexander said. “Step one is, have a clear policy. Step two is, make sure you’re being accountable to families who are directly impacted. Is it working? How do we make it better?”
Rodríguez said since the San Francisco policy passed, several other mothers have thanked her. She said she would love for SB 445 to pass so parents in other districts can also benefit.
“So many children in many different places, many different schools, are not receiving the support they deserve, and their parents have to battle to get an evaluation and to get documents translated, and they find it really hard,” Rodríguez said. “It’s a really, really long document, and it’s a long process. And if it’s in our language, then it will be much easier for us parents to process and understand the document and the evaluation given to our children.”
We explain the meaning of ‘hallucinate’, and why the Cambridge Dictionary chose it as its Word of the Year.
Hallucinating ‘false information’
“To seem to see, hear, feel, or smell something that does not exist, usually because of a health condition or because you have taken a drug.” This used to be the definition of ‘hallucinate’ in the Cambridge Dictionary prior to 2023.
Advertisement
But in the year when generative AI tools such as ChatGPT and Bard, which use large language models (LLMs), captured the world’s imagination, ‘hallucinate’ has taken on additional meaning. The Cambridge Dictionary, this year, made the following addition to its definition: “When an artificial intelligence hallucinates, it produces false information.”
As we live through the nascency of generative AI with all its deficiencies and limitations, hallucinations are all too common. Sometimes utterly nonsensical while seemingly plausible other times, they differ in form. But at their core, they are falsehoods touted confidently by AIs while processing any prompt.
Why AIs hallucinate
Generative AI takes actions based on past data. When given a prompt — a piece of text, an image, or even a piece of computer code — the AI generates an appropriate response based on the information it has been trained on. Popular tools such as ChatGPT and Bard use LLMs, which learn from ginormous data. Effectively, they (attempt to) recreate human thought and linguistic expression by ‘learning’ from millions of (human-created) sources.
“At their best, large language models can only be as reliable as their training data,” Wendalyn Nichols, Cambridge Dictionary’s Publishing Manager, said in a statement. “AIs are fantastic at churning through huge amounts of data to extract specific information and consolidate it. But the more original you ask them to be, the likelier they are to go astray,” she said.
Most Read
1 South Africa vs Australia Highlights, World Cup 2023 Semi-Final: Australia to play the WC final against India after close shave against South Africa at Eden
2 Tiger 3 box office collection day 4: Salman Khan-starrer records 50% dip, earns Rs 169.5 cr in India
Advertisement
AIs can learn from factually inaccurate sources, or produce inaccuracies themselves while processing the information. Either way, hallucinations have some real-world consequences. For instance, a US law firm used ChatGPT for legal research, which led to fictitious cases being cited in court. The judge fined the firm a sum of $ 5,000 for the mistake.
Why Cambridge Dictionary chose ‘hallucinate’
The Cambridge Dictionary Word of the Year has been published every year since 2015. A team makes the decision based on data on words that saw popular usage in the year and their cultural salience.
“The Cambridge Dictionary team chose hallucinate as its Word of the Year 2023 as it recognised that the new meaning gets to the heart of why people are talking about AI,” the company’s statement announcing the decision on Wednesday read. “Generative AI is a powerful tool but one we’re all still learning how to interact with safely and effectively — this means being aware of both its potential strengths and its current weaknesses,” it said.
What word defined the past year for you? Find out if your guess matches the actual word of the year from two prominent dictionaries.
If there was one word that described the year you just had, what would it be? Maybe it’s one of the 690 new words and phrases Merriam-Webster just added to its dictionary—like beastmode for the workout routine you’ve kept up with since New Year’s, or chef’s kiss after you finally figured out how to make the perfect chocolate chip cookie. Or maybe it’s something you heard repeatedly in conversation, like a cultural trend, food, acronym or even a concept like the Roman Empire. (Yes, the actual Roman Empire.)
Whatever your personal word of the year may be, some words tend to be more broadly significant and influential, according to the world’s most prominent dictionaries. These are the terms that might refer to a cultural zeitgeist, a controversy or our larger thoughts (and often anxieties) about the world. Cambridge Dictionary and Collins Dictionary recently revealed their words of the year for 2023, and they have all these components. Can you guess what they are? Read on to find out.
Get Reader’s Digest’s Read Up newsletterfor more news, humor, cleaning, travel, tech and fun facts all week long.
Cambridge Dictionary’s word of the year
RD.com
Cambridge Dictionary’s pick for 2023 is—drum roll, please—hallucinate. You might be scratching your head right now, since this isn’t a new word, per se. Of course, the common definition of hallucinate is “to seem to see, hear, feel or smell something that does not exist, usually because of a health condition or because you have taken a drug.” In 2023, however, to hallucinate can mean something different, thanks to AI.
According to Cambridge‘s alternate definition of hallucinating, “when an artificial intelligence (= a computer system that has some of the qualities that the human brain has, such as the ability to produce language in a way that seems human) hallucinates, it produces false information.” Sure, using AI can be fun for creating dog selfies or could even help you land a job, but it’s prone to producing misleading or made-up facts—or “hallucinating.”
Which means we’ll likely be dealing with more misinformation, at least in the near future. “The fact that AIs can ‘hallucinate’ reminds us that humans still need to bring their critical-thinking skills to the use of these tools,” notes Wendalyn Nichols, Cambridge Dictionary‘s publishing manager. Translation: Don’t believe everything you read—or that AI tells you!
Collins Dictionary’s word of the year
RD.com
Those dictionary people are apparently (ahem) all on the same page—or at least the editors at Cambridge and Collins are. Similar to Cambridge‘s choice, Collins Dictionary‘s word of the year is the broader term AI. Collins chose this term because it is “considered to be the next great technological revolution,” “has seen rapid development” and “has been much talked about in 2023.”
ChatGPT was released in late 2022, with companies attempting to use it to cut costs, worrying employees that it would replace their jobs. AI pioneers and creators began expressing concerns that AI could be “dangerous” and manipulated by “bad actors.” President Biden even issued an executive order on “safe, secure and trustworthy artificial intelligence.”
And then, of course, there were the months-long writers and actors strikes, much of which hinged on the potential use of AI. While humans eventually won in both scenarios—with the new contracts stating that AI is not allowed to write or rewrite content or use an actor’s likeness in a way that the actor didn’t originally agree to—this is likely the first battle of many to come, in a variety of industries.
