Here’s an interesting COVID infection control animation from Sanford Medicine that tells patients and caretakers what they need to know without the use of words. Well worth a look for both the information conveyed, and for alternate methods of presenting information to people who speak many of the languages of the world.

Among the dislocations of the COVID-19 pandemic is the effect on health system interpreters, who are discovering new complexities in the already fraught world of medical interpretation. As revealed in the New York Times story, When Coronavirus Care Gets Lost in Translation, interpreters are increasingly called on to work over the phone in consideration of the lack of protective equipment and the risk of infection. This already difficult situation is rendered more complex by the distance providers attempt to put between themselves and their patients, the muffling effect of masks, and the hiss of oxygen tanks.

“Someone’s oxygen can be dropping and I have to get an interpreter on the phone, wait on hold, put in an access code, tell them where I am,” one Boston doctor said. “It’s hard for the patient. Imagine you’re in a loud room with a mask blowing oxygen in your face at 15 liters a minute and you feel crummy. You can’t comprehend things as much.”

The piece explores this as another health care disparity issue that results in poorer outcomes for non-white patients. Patient advocate Vonessa Costa told the Times, “A pandemic is not the time to build working systems.” She worried about the millions of limited English patients across the country who are now or will be struggling for access to care.

Consider the poem, A une Damoyselle Malade (To a Sick Young Lady), written in 1537 by Clemont Marot for the daughter of a queen. Twenty eight lines, three syllables per line. How complicated can it be to translate from French to English?

The answer: plenty complicated. Douglas Hofstadter, author of the wildly acclaimed 1999 non-fiction title, Godel, Escher, Bach: An Eternal Golden Braid, takes on this question in a podcast featured on the NPR program, Radiolab.

“What kind of of crazy things can happen when you translate crazy texts,” Hofstadter asks. He set off of a project to translate the poem himself, then conscripted 60 others to take on the job. Soon he found himself embroiled in questions regarding tone and form — how do you preserve the feel of the poem, the lightness and good cheer — amid the myriad possibilities offered by the English language?

While not immediately analogous to the problems faced by medical interpreters and translators, this is another, interesting take on the complexity of translation. “What is translation?” Hofstadter finally asks. “And can it be done?”

Hofstadter’s interview, titled 100 Flowers, is part of an hour-plus presentation that features seven other related segments.

Yet another sign that the future is catching up with us fast: the availability of interpretation devices that fit in your ear and instantly render foreign languages into your own.

A recent article in the New York Times, The Latest in Translation Devices, describes two additions to what is likely to be a booming market — the WT2 Plus Ear to Ear AI Translator Earbuds from Timekettle, and the  “Ambassador” from Wavery Labs. Both rely on a pair of wireless earpieces that are synced to a single smartphone connected to Wi-Fi or cellular data.

The Times article quotes Graham Neubig, an assistant professor at the Language Technologies Institute of Carnegie Mellon University and an expert in machine learning and natural language processing, who says, “These devices ‘bring us a bit closer to being able to travel to places in the world where people speak different languages and communicate smoothly with those who are living there.'” At the same time, the Times story makes it clear that human translators and interpreters will not be replaced by this generation of devices.

Read the complete Times piece here.

Care to geek out on the technology behind this leap in the processing and rendering of speech and writing? Check out a story in the New Yorker that delves into an explanation of machine learning while attempting to answer the question, Can a machine learn to write for the New Yorker? (The answer: not quite yet.)

Here’s another item to stuff into your Brave New World file — a hand-held translation device that, according to Bloomberg News, is taking Japan by storm and is also now available in the US. (“Tourists are fueling a boom in personal translation devices”)

The $299 product can be worn around the neck, and is capable of translating phrases to and from 74 languages. With a tourist boom that brings 31 million visitors to Japan each year, shop owners looking for additional profit need a way to communicate with customers. The Pocketalk device allows store keepers to capture a spoken phrase from foreign customers, which it then translates into Japanese (or, for that matter, English or any of the other supported languages).

Does it work? Amazon.com reviews paint a mixed picture, with some users complaining that the device is too slow and cumbersome, while others describe it as hugely helpful. Machines aren’t ready to replace the subtle understanding and nuance that human interpreters bring to complex medical encounters. But if all you really want is a platter of sushi in Tokyo…?

Here’s an inside look at a skilled interpreter’s complicated work. This video focuses on interpreting in a diplomatic environment as it explains the advantages and pitfalls in simultaneous and consecutive interpretation — but the lessons are broadly applicable to the medical world as well. Along the way you’ll learn the difference between chuchotage and décalage, and also find out why all too often telling a joke through an interpreter is a dumb idea.

Despite the advances in machine translation that have moved it from the ridiculous to the potentially useful, there remain glitches — particularly in the area of assigning gender, as pointed out in an article in Slator: Language Industry Intelligence.

A difficult trouble spot is in translations that go from English, which does not assign gender to nouns, to languages such as Spanish or French, which do.

Quoting a Dublin City University researcher, Eva Vanmassenhove, the Slator article reports, “In French and Spanish, for instance, ‘I am a nurse’ will still be given feminine translations while ‘I am a surgeon’ will result in masculine ones.

“Even more problematic is the following set of translations: ‘I am beautiful’ is translated into the male form in Spanish. However, ‘I am a beautiful surgeon’ is translated into a female form,” Vanmassenhove said.”

Read more about another complexity of the modern moment in “He Said, She Said: Addressing Gender in Neural Machine Translation.”