Saturday, October 28, 2023

Wanted: New definitions of old words - Moneycontrol - Dictionary

Each entry in Gustav Flaubert’s (Madame Bovary) dictionary takes a common phrase and exposes the inanity of its usage. Under “duties”, he writes: “Demand them of others, free oneself from the same.” (Image via Pexels/Pixabay)

Each entry in Gustav Flaubert’s (Madame Bovary) dictionary takes a common phrase and exposes the inanity of its usage. Under “duties”, he writes: “Demand them of others, free oneself from the same.” (Representational image via Pexels/Pixabay)

Words and phrases are slippery things. Nowadays, many are used in baffling ways: for example, “self-defence”, “rules-based international order”, “civilization” and “democracy”. In Lewis Carrol’s Alice in Wonderland, Humpty Dumpty may have scornfully proclaimed that when he used a word, it meant just what he chose it to mean—but, as Alice replied: “The question is whether you can make words mean so many different things.”

Over the years, some writers have taken pains to point out how the meanings of words can shift to suit different ends. Take Gustave Flaubert’s Dictionary of Received Ideas, published posthumously in 1911, which mocked the clichés and stereotypes used by French society under Napoleon III. In these notes and jottings, Flaubert satirized shallow, unthinking attitudes, especially of the bourgeois, which he had earlier made evident in Madame Bovary.

Each entry in Flaubert’s dictionary takes a common phrase and exposes the inanity of its usage. Under “duties”, he writes: “Demand them of others, free oneself from the same. Others have them towards us, but we have none towards them.” For “era”, he says: “Thunder against it. Complain that it lacks poetry. Call it an age of transition, of decadence.” Other aspects of his time that needed to be thundered against were newspapers, war and feudalism.

In the same vein, gibberish is merely “a foreign people’s way of speaking,” and imbeciles are “those who think differently from oneself”. What about censorship? Well, “it has its uses, say what you will”. As for illusions: “Pretend to have many, complain about having lost them.” When it comes to people and professions, “all journalists are ideologues,” an Orientalist is “a man who is well-travelled,” and imperialists are “honest, polite, peaceful, distinguished people”. But of course.

A few years before the publication of Flaubert’s gibes, there was Ambrose Bierce’s The Devil’s Dictionary, originally titled The Cynic’s Word Book, a collection of newspaper columns. Bierce’s aim, allied to Flaubert’s, was to provide a subversive take on the English language that highlighted the absurdities and contradictions in communication, politics, and society.

Bierce’s definitions are sharp-tongued, and can make one wince with the light they shine on everyday insincerity. He is an equal-opportunity offender, and no respecter of faiths. For him, a dictionary is “a malevolent literary device for cramping the growth of a language” but his own lexicon “is a most useful work”.

In particular, his observations on politics and governance have stood the test of time. An alliance is, in international affairs, “the union of two thieves who have their hands so deeply inserted in each other's pockets that they cannot separately plunder a third”. Politics itself is defined as “a strife of interests masquerading as a contest of principles”. It is “the conduct of public affairs for private advantage”, while diplomacy is “the patriotic art of lying for one’s country”.

In the same vein, patriots are those “to whom the interests of a part seem superior to those of the whole”. They are “the dupe of statesmen and the tool of conquerors”. As for patriotism: “Combustible rubbish ready to the torch of anyone ambitious to illuminate his name.”

Such are the unedited thoughts of one for whom history is “an account mostly false, of events mostly unimportant, which are brought about by rulers mostly knaves, and soldiers mostly fools”. If you disagree, you may be an idiot: “A member of a large and powerful tribe whose influence in human affairs has always been dominant and controlling.”

If all else fails, there’s always gunpowder, which is “an agency employed by civilised nations for the settlement of disputes which might become troublesome if left unadjusted”. War, as Clausewitz said, is the continuation of policy by other means. This can lead to happiness: “an agreeable sensation arising from contemplating the misery of another.”

Many of these definitions still resonate because meanings change but human folly remains unchanging. Teju Cole has rightly pointed out in reference to such dictionaries that we use clichés as crutches, propping up our lazy, prejudiced, or hypocritical opinions as if they were profound or fresh insights. It’s time we exposed this duplicity with a resounding “Bravo!”

Adblock test (Why?)

Hundreds of Alachua County parents urge translation services from school district - WCJB - Translation

GAINESVILLE, Fla. (WCJB) - Hundreds of parents in Alachua County are demanding the school district to step up efforts in providing information in different languages.

“It’s a right that any parent, any student in the school district can request,” shared Adriana Menendez with Rural Women’s Health Project. “They have the right to receive translation services. They should be able to receive information in their language.”

Menendez advocates for immigrants in the county and said she’s witnesses the problem first-hand.

“One time an only Spanish-speaking parent received a ‘no school’ notice in English,” stated Menendez. “They dropped off their child at school. When their child opened the classroom door, there was no teacher inside. There were no students inside.”

Advocates, like Ethan Maia with Gainesville Immigration Neighbor Inclusion Initiative (GINI) said the school district is not doing enough to bridge the language barrier between parents and staff.

“Many of these families are just left out completely in the dark,” shared Maia. “Or if they’re trying to reach out to the schools, they are unable to get an interpreter in their language so they can actually effectively communicate with their child’s school.”

Menendez told TV20 more than 600 parents signed a petition which requests additional translation services. Meanwhile Alachua County school district officials said they’re working on new initiatives.

“We instituted the Language Line program that allows us to communicate with parents by phone or through zoom, in 200 plus languages including American sign language so we’re getting a lot of use out of that,” shared ACPS spokesperson Jackie Johnson.

Johnson said the school district hired a Spanish translator in August 2022. According to Skyward, there are more than 4000 students who come from Spanish-speaking households in the county.

“Sometimes we ask for a translation service or for someone to translate and we have to wait a long time,” shared parent Marvin Ramos. “Assistance in other languages would also be great because people speak several languages, not just Spanish and English.”

Johnson claims district officials are working to break language and cultural barriers.

TRENDING STORY: MCSO releases body camera video of deputy-involved shooting

Click here to subscribe to our newsletter.

Adblock test (Why?)

Friday, October 27, 2023

The White House and Google launch a new virtual tour with audio captions, Spanish translation - ABC News - Translation

WASHINGTON -- Can't come to Washington? Couldn't get a ticket to tour the White House? Don't worry.

The White House, Google Maps and Google Arts & Culture launched a new virtual tour of the famous mansion on Friday, which is also National Civics Day.

With a computer or smartphone, users will be able to spend time zooming in on all of the rooms that they would have seen had they been able to go on an in-person tour.

The updated virtual tour is part of a desire by first lady Jill Biden to make the White House accessible to as many people as possible. Biden, a longtime community college professor, hopes teachers use it to educate students about the White House and its history, said Elizabeth Alexander, her spokesperson.

“Not everyone can make the trip to Washington, D.C., to tour the White House, so she's bringing the White House to them,” Alexander said.

Biden traveled to Philadelphia on Friday for a National Civics Day event hosted by Nickelodeon, ATTN: and iCivics, where they announced “Well Versed,” a new short-form series that uses animation and music to help teach children about democracy and the Bill of Rights.

She talked about the Constitutional Convention held at Philadelphia's Independence Hall in 1787, where the founders of the United States created a government in which power rested with the people, not with kings and queens.

“That’s still how our country works, and it’s one of the things that make it so special," Biden said. "And when we understand civics, how our government works and how to hold it accountable, we are able to help each other and make our country the best it can be.”

The virtual tour is the first Google virtual tour of the White House to include audio captions for people with disabilities. The captions are narrated by White House social secretary Carlos Elizondo and pop up on the screen to offer viewers historical information on each of the rooms.

It is also Google's first virtual tour of the White House to have Spanish translation, and feature the official portraits of former President Barack Obama and former first lady Michelle Obama.

The tour opens with a brief video of President Joe Biden and the first lady welcoming visitors, the same message that plays at the White House Visitors Center for those who visit in person.

Google Street View technology was used to capture the imagery, starting at the East Wing Entrance and moving through all rooms on the public tour route, including the library, the China Room, the Green, Blue and Red rooms, the East Room and the State Dining Room.

The tour was created using Google Arts & Culture’s storytelling tool.

Ben Gomes, senior vice president of learning and sustainability at Google, said the mission of its arts and culture division is to open the world's culture to people everywhere.

The tour is available on the White House website, as well as on Google Maps and the Google Arts & Culture page.

Adblock test (Why?)

Microsoft Says the ComSL Model Outperforms Other Models in Speech Translation - Slator - Translation

On October 14, 2023, researchers at Microsoft Cloud and AI, Microsoft Research Asia, and Shanghai Jiao Tong University published updated results for the capabilities of ComSL (Composite Speech-Language Model), a speech-language model originally introduced in a paper in May 2023.

According to the researchers, the ComSL model is based on public pretrained speech-only (audio data) and language-only (text data) models and has been optimized for spoken language tasks by integrating both modalities into its training.

The main differentiator of the ComSL model, explained the researchers, is that it outperforms the results achieved through “end-to-end modeling,” the most widely used training methodology thus far. End-to-end modeling uses audio and text data separately even if, the researchers say, they “may not be optimal for each other.”

In the composite model, the researchers obtained a simpler cross-modality learning that uses speech-text mapping/matching. The training allows the model to perform better and does not require any force-aligned speech and text.

For their methodology, the researchers applied machine translation (MT) and automated speech recognition (ASR) as what they call “auxiliary tasks” in a multi-task learning mode during the optimization of the end-to-end speech translation (ST) model.

10 LLM Use Cases (Main Title)

Slator Pro Guide: Translation AI

The Slator Pro Guide presents 10 new and impactful ways that LLMs can be used to enhance translation workflows.

Multi-task learning (MTL) mode implies “sharing common knowledge among different tasks” so that the MT task can guide the ST task. However, the researchers stated that, because of the mismatch between speech and text modalities, the guidance was not as effective.

The ComSL model was trained with existing, fine-tuned models, including speech-only input and text-only input, as well as with ST, ASR, and MT as tasks and a “cross-modality learning (CML)” approach based on paired speech-text input instead of forced-alignment. 

The training steps consisted of fine-tuning the language model (with all the paired text data), multi-task learning (the tasks were ST, MT, ASR, and CML), regularization on the MT output (fine-tuning with MT tasks), and freezing speech encoder (retaining speech representations at the start of fine-tuning).

400 hours of English

The experiments in this study involved the CoVoST 2 dataset, which comprises translations from 21 languages into English and from English into 15 languages, and approximately 400 hours of English recordings and 900 hours of recordings from 21 additional languages. 

The researchers focused mainly on the non-English language into English speech translation, measuring performance with BLEU scores and the CoVoST 2 testing set. The models utilized as the baseline were Whisper and mBART-50, themselves fine-tuned with CoVoST 2.

The composite model was found to outperform the base speech model (Whisper) and the combination of speech and language models (Whisper+mBART). The incorporation of ST data contributed to a high score on the CoVoST2 testing set, and the composite model was also evaluated on speech-to-text translation tasks with better results than those known for the end-to-end modeling that includes the same tasks of ST, ASR, and MT.

Adblock test (Why?)

How Effective Are Large Language Models in Low-Resource Language Translation - Slator - Translation

Large language models (LLMs), such as ChatGPT, have shown remarkable capabilities in performing a range of language tasks, including machine translation (MT). But how effective are they when it comes to low-resource languages (LRLs)?

A research paper published on September 14, 2023, delves into the translation prowess of ChatGPT and other LLMs across a diverse set of 204 languages, encompassing both high- and low-resource languages. According to the authors, this is “the first experimental evidence for an expansive set of 204 languages.”

Nathaniel R. Robinson, Perez Ogayo, David R. Mortensen, and Graham Neubig from Carnegie Mellon University underscored the need for such an investigation, noting that there exists a wide variety of languages for which recent LLM MT performance has never been evaluated. As a result, it is difficult for speakers of the world’s diverse languages to know how and whether they can use LLMs for their linguistic needs.

In addition, the authors emphasized that “the majority of LRLs are largely neglected in language technologies” in general with current MT systems either performing poorly on them or not including them at all. “Some commercial systems like Google Translate support a number of LRLs, but many systems do not support any,” they said.

The authors pointed out that their work differs from existing studies since the focus here is on end users. The inclusion of a remarkable 204 languages, which incorporates 168 LRLs, underscores the commitment to addressing the diverse needs of LRL communities, which are frequently overlooked in the discourse on language technology. “We include more languages than any existing work […] to address the needs of various LRL communities,” they explained.

10 LLM Use Cases (Main Title)

Slator Pro Guide: Translation AI

The Slator Pro Guide presents 10 new and impactful ways that LLMs can be used to enhance translation workflows.

To conduct their research, the team used data from FLORES-200 (an evaluation benchmark) and queried the OpenAI API to translate their test set from English into the target languages. 

They evaluated ChatGPT’s MT performance across the entire language set and compared it with NLLB-MOE as their baseline, as it is the current state-of-the-art open-source MT model with wide language coverage. Comparative evaluations were also carried out against results from subsets of selected languages using Google Translate and GPT-4.

In their exploration of MT prompts, they employed both zero- and five-shot approaches for ChatGPT MT. The evaluation metrics, spBLEU and chrF2++, provided a robust basis for assessing the outputs.

The results suggest that while ChatGPT models approach or even surpass the performance of traditional MT models for some high-resource languages, they consistently lag for LRLs. Notably, African languages emerge as a particular challenge, with ChatGPT underperforming traditional MT in a substantial 84.1% of the languages studied.

Language Resources and Costs

The researchers also examined language features, including language resources, language family, and script, to assess the effectiveness of LLMs. 

This analysis aimed to uncover trends that could guide end users in selecting the most appropriate MT system for their specific language. “Analyzing this may reveal trends helpful to end users deciding which MT system to use, especially if their language is not represented here but shares some of the features we consider,” they said.

According to the authors, a language’s resource level is the most important feature in predicting ChatGPT’s MT effectiveness, while script is the least important.

The authors stressed financial aspects as well, particularly as it pertains to LLM users. “We evaluate monetary costs, since they are a concern for LLM users,” the authors said. Few-shot prompts, despite their potential for modest improvements in translation quality, come at a higher cost due to charges for both input and output tokens.

The authors emphasized that they want to help end users of various language communities know how and when to use LLM MT. “We expect that our contributions may benefit both direct end users, such as LRL speakers in need of translation, and indirect users, such as researchers of LRL translation considering ChatGPT to enhance specialized MT systems,” they concluded.

Adblock test (Why?)

Thursday, October 26, 2023

Interpreter says Puska asked for confession translation - RTE.ie - Translation

An interpreter has told the Central Criminal Court that the man accused of the murder of schoolteacher Ashling Murphy asked him to translate his confession to gardaí, two days after she was killed.

Miroslav Sedlacek was giving evidence at the trial of 33-year-old Jozef Puska who has pleaded not guilty to the murder of Ms Murphy in January 2022.

Miroslav Sedlacek is originally from the Czech Republic and provides translation services in German, Czech and Slovak.

On 14 January 2022, he provided translation services in Slovak twice to gardaí in St James' Hospital in Dublin, on a phone line.

He told the court the second conversation took place at around 6pm on that evening, and lasted around 20 minutes.

He said the conversation began with gardaí telling Jozef Puska about the search warrant they had and explaining that his personal belongings would have to be seized for an investigation into a murder in Tullamore.

He told the court Mr Puska wanted to know how this was related to him and wanted to know if he was a suspect. Mr Sedlacek said gardaí told him he was a person of interest and explained what this meant.

Mr Sedlacek said he remembered very well what followed after this. He said it was at this point that Mr Puska asked him personally to translate his confession.

He said Mr Puska asked him to translate accurately and exactly what he was saying. He said Mr Puska told him to tell the gardaí that he did it, that he killed her and that he did not do it intentionally.

Mr Sedlacek said this was still between him and Mr Puska before he had the chance to translate – it was quite spontaneous he said, everything came quickly.

He said Mr Puska said he did not want to do it, that he was very sorry that he did it and that it happened. Mr Sedlacek said he translated to gardaí word for word and gardaí cautioned Mr Puska. He said he translated the caution and Mr Puska said he understood.

Mr Sedlacek said Mr Puska then started asking some questions.

He said Mr Puska was very concerned about the safety of his family. His first concern was whether or not his family members’ names would go public. Gardaí said his own name would go public.

He also asked if there was any possibility the girl’s family would like to take any revenge on his own family for what he had done to her. He said gardaí explained Ms Murphy’s family would certainly not take revenge on his family.

Mr Sedlacek said Mr Puska’s voice was very different from the first conversation he had with him earlier on the same day. He said he was quite emotional and his voice was trembling, adding his sentences were quite disjointed. He said he supposed this was as a result of the situation he was in.

He said Mr Puska wanted to stress that he did not do anything intentionally.

He said the garda then told him that Mr Puska was not feeling well and they would have to end the call.

He said Mr Puska asked what would happen next and the garda explained that when he recovered he would be brought to Tullamore garda station and would be interviewed there.

Mr Sedlacek said he would describe Mr Puska as being in very low spirits after the confession. "I would even say desperate," he told the court.

Earlier, the site nurse manager at St James’ Hospital, Roz Gillen, told the court she had been approached by Detective Sergeant Pamela Nugent on the evening of 14 January. The garda had a copy of a search warrant and Ms Gillen decided to move Mr Puska to a single room.

Under cross examination from defence counsel Michael Bowman, Ms Gillen said there was never any request by gardaí to speak to a treating doctor. She was not asked to refer to his medical notes and had no understanding of Mr Puska’s state of mind or medical circumstances.

She agreed she had no function in determining the fitness of someone to deal with gardaí.

Asked if a request to deal with a treating doctor could have been accommodated, she said she did not know if a doctor would have been there as it was a Friday evening.

Adblock test (Why?)

The rise and inevitable fall of Joy Pocket Dictionary - The Business Standard - Dictionary

The red cover caught my attention on a recent afternoon stroll down a footpath at Old Paltan. It was a small Joy Pocket Dictionary that stood out from the hundreds of books by celebrated and amateur writers. 

I instantly recognised it. In my school days, I always carried the dictionary with me. I studied new words and found my favourite ones. Truth be told, I often forgot what I learned, so I would then try to memorise the word's meanings again. This kept on for a few years.       

There was a time when the Joy Pocket Dictionary was ubiquitous across the country. School and college-going students who wanted to improve their English language skills would always carry it. 

The Joy Dictionary had to compete with its rival Indian AT Dev's pocket-sized dictionary and lived through the tough competition of the dictionary business in the 1990s and 2000s. 

However, it was the emergence of websites, and later mobile phone applications, which ultimately proved to be the final nail in Joy Pocket Dictionary's coffin. But before it met its demise and became a thing of nostalgia, the dictionary saw outstanding business.

The highs and lows  

At the height of its popularity between 1990 to 2000, the sale of a single category of dictionary reached 10,000 copies per month. Sometimes, special discount periods like Pahela Baishakh saw higher sales. 

"We would jointly make efforts to scale up the business," said Shahid Hasan Tarafder, the owner of publishing company Gyankosh Prokashoni, adding, "The binding and the cover were also attractive." 

At the time, there were around 10 product lines including pocket dictionaries, learner's dictionaries, advanced learner's dictionaries, and Joy concise dictionaries. The company used to publish English language learning books like Six-in-One and Three-in-One. There were some religious books too.

In 1988, Shahid became the sole agent for Joy Dictionaries in Dhaka city.

He bought the copyrights in 2006. By then, the internet had already reached city homes, businesses, offices and cyber cafes in district towns, but people were not quite accustomed to it. Also, there was the factor of regular accessibility to the internet. As a result, the Joy Dictionary continued to enjoy massive popularity. 

Shahid said that Joy's pocket dictionaries, as well as the medium-sized Joy Advanced Pocket Dictionary, were sold at the same pace. The other Joy dictionaries include different versions — English-to-Bangla, Bangla-to-English and Bangla-to-Bangla. The most popular of them is still the English-to-Bangla dictionary. 

Gyankosh Prokashoni saw a boom in Joy Dictionary sales for approximately 10 years. 

However, a gradual decrease in sales started to emerge. During 2015-16 and due to the emergence of mobile phone apps, sales started to take a nosedive. Fast forward to 2023, Shahid said that the number of sales of a single dictionary has now come down to 500 copies per year. 

In 2006, the price of Joy Pocket Dictionary was more or less Tk20. Now the wholesale price is Tk40. 

Every year, Shahid's publishing company publishes around 10,000 copies of Joy dictionaries to run the whole year. The first edition of the dictionary came out in 1985. Another edition came out in 1990. But the dictionary was reprinted in 2023. 

Additionally, Shahid said more than 100 words have been added to the dictionary in the last decade by the editors.  

The rise of Joy Dictionary 

SK Ahmed was the original publisher of the Joy Pocket Dictionary. In the mid-1980s, publishing company Joy Books International started to publish the dictionary. 

"He [SK Ahmed] produced the dictionary and I would distribute the dictionaries across the country," said Shahid, now a 67-year-old man.  

"In the 2000s, at one point, SK Ahmed lost interest in the book business. He proposed that I buy out his company," recalls Shahid. "For Tk50 lakh in 2006."    

"He is one of my distant relatives, and I knew the ins and outs of the market of the Joy Dictionary," Shahid added.

He had another reason to buy out Joy Dictionary. Gyankosh, Shahid's stationary shop which started in 1980 mainly with academic textbooks, became popular with customers because of this dictionary.   

SK Ahmed Publishing Company was the first local private book publishing company to publish pocket-size and medium-size dictionaries in Bangladesh, Shahid said. 

"[And] the quality of the dictionary was always good," said Shahid. He said that many Bangladeshi publishers later took the initiative and published dictionaries but failed to replicate Joy Dictionary's success. 

Shahid also recounted how SK Ahmed had a printing press in the New Market area. "This man had the capacity to do something innovative. The Indian imported dictionaries received a blow because of the Joy Pocket Dictionary for its quality," said Shahid.  

At that time, some Indian pocket dictionaries of AT Dev reached the market in Bangladesh. But Joy Dictionary put a stop to those imports. 

However, the book-size Samsad Dictionary continued to reach Bangladesh, and to date, some still do. 

Joy turns to despair 

The making of a pocket-size dictionary is difficult. The bookbinders who once used to bind pocket-size dictionaries now show no interest. Shahid explained that the price of the paper has also contributed to the near-demise of the small-size dictionary. The profit margin has fallen significantly. 

In 2018, the price of one rim of double-demy paper stood at Tk800 to Tk900. Now one rim of the double-demy paper is Tk1,700 to Tk1,800. The wages of the bookbinders have also gone up. 

"The profits are not even half of what we used to make in the past," said Shahid.

But it is the weight of the disappearing interest in hard copy dictionaries that decided Joy Dictionary's demise. "In the past people would buy a dictionary with enthusiasm. That enthusiasm has gone away," explained Shahid.    

Currently, they sell Joy Pocket Dictionary, (English to Bengali), Joy Pocket Dictionary (Bengali to English), Joy Nobo Obidhan, Joy Advanced Pocket Dictionary (Bengali to English), Joy Standard Pocket Dictionary, Joy Shabdha Shanchayeta. 

A new dictionary on the cards?

Gyankosh Prokashoni has taken the initiative to publish a book-size Joy Advanced Learners Dictionary this year. 

"Many publishers have a book-size dictionary. As businessmen, we have to always keep up with the competition," said Shahid, adding, "We have made some progress."    

Asked about a plan to make a mobile application for Joy Dictionary, he said, "My son Abdul Wasif has gotten involved in the business and he will decide on the matter. He has some plans for something like that." 

Gyankosh Prokashoni has not changed the logo or colour of Joy Dictionary till now, having only added their names as the publisher. They even kept the name of the former publisher. The dictionary was edited by SK Ahmed in collaboration with experienced professors and headmasters. 

"I kept the name because it is a matter of courtesy and honour," said Shahid. 

Adblock test (Why?)