Much to the surprise of Vocabulary.com, a sound clip from their site has become somewhat of an overnight Internet sensation. Have you heard it?
What do you hear? “Yanny” or “Laurel”?
With just under 50 percent of listeners hearing “yanny” and just over 50 percent hearing “laurel”, this phenomenon has sparked quite the debate on what is really being said. So, how did this madness begin, and what does it mean if anything for the language services industry? We’re glad you asked, so let’s get right to it, shall we?
It really is amazing how some Internet sensations get their start. In the case of the “yanny vs laurel” madness, it all started at a high school in Georgia. A freshman by the name of Katie Hetzel was studying for her world literature class, and lo and behold, “laurel”, was one of her vocabulary words:
A laurel is a wreath worn on the head, usually as a symbol of victory. If you see an image of Julius Caesar, chances are he’s wearing a laurel.
A laurel is a type of wreath — circular, made from leaves and branches — worn on the head in ancient times. The laurel is a symbol of victory that lives on in the phrase “Resting on one’s laurels.” When you rest on your laurels, you’re happy with previous successes but not doing much to continue succeeding. You’ve gotten lazy and complacent.
Listen to the audio clip here – Source: Vocabulary.com
However, Ms. Hetzel didn’t hear the word, “laurel” when she played the sound bit – she heard, “yanny.” Intrigued, she decided to ask some friends what they heard, and the answers were mixed. As is customary for today’s American teenagers, this intrigue moved swiftly to the Internet. First posted on Instagram, the curious sound bit was then picked up by another student who started to poll listeners on what they heard. It then traveled to Reddit, courtesy of yet another fellow student where it very quickly went viral. Oh, those high schoolers!
You may be surprised that, as machine-like as it sounds, the clip was actually recorded by a human voice – and not any ordinary voice but the voice of a trained opera singer. It has become the custom at Vocabulary.com, to hire individuals who have a strong background in (and who are highly skilled at) using the international phonetic alphabet (IPA). Since opera singers sing in languages that they don’t necessarily speak, they often have a very firm grasp of IPA. Vocabulary.com uses high-quality microphones but in a DIY sound booth. As words appear on the screen, opera singers pronounce it – again, again, and again – until it is just right. And that, in a nutshell – well, actually, in a sound booth – is how the now fanatically popular laurel sound clip was born.
It all circles back to the brain. You know, that 3-pound powerhouse that sits atop our shoulders? Everything we do, say, hear, feel, see, and touch becomes imprinted in our brains to one degree or another. We then subconsciously call on these experiences when we are presented with strange or unique sensations. When we listen to a newly presented sound, in this case, “laurel”, if the brain cannot clearly decipher what is being said, it will pull from prior experiences to try and fill in the missing gaps.
It’s also about frequencies. According to Professor Brad Story who works for the University of Arizona’s Speech Acoustics and Physiology Lab, the shape of a human’s throat and mouth create a variety of frequencies, the lowest of which is used to encode language as a sound wave. The bottom line? Humans do actually hear differently based on our experiences and our physiology.
But we must also consider what device people are using to hear this audio clip. Are they listening to it on their laptops, their mobile devices, or their tablet or iPad? Are they listening to the sound and watching a screen? What is the quality of the recording, and what visuals might be affecting their brain’s interpretation of the sound?
There is a lot more that goes into what we hear than mere sound itself.
So, what does this mean for those of us obsessed with language, and what does this mean to those of you who require audio localization? Lots.
If you are gearing up to hire an audio localization vendor, chances are you already have a script that represents your brand, your voice, and your image, right? And although you know exactly how you want that message to be shared with prospective clients and customers, how can you guarantee they will hear it the way you do?
In the United States alone, your message can potentially be received by millions of foreign-born prospective clients and customers. If you fail to effectively localize, you could be missing out.
The population of foreign-born US residents from other Latin American countries such as the Caribbean, South America, and Central America, as well as foreign-born residents from the Middle East have increased by factors of 10 or more since the early 1960s.
Start by either gathering voice samples or having the audio localization vendor present you with a number of samples from their repertoire. What age, gender, and style of voice will best capture your message and resonate with your target audience? Get as many samples as the vendor can muster up and as many as your budget will allow. The more samples, the more likely you are to find that one dream voice that says it all.
It’s not just about the voice. If you are targeting regions beyond your local audience, having your script professionally translated and localized ensures that your message will be culturally and linguistically sensitive to all prospective clients and customers. And, if the professionals you hire to localize and translate are really on top of their game, the final product won’t just be linguistically and culturally sensitive – it will also be highly effective in those markets.
You now have the voice and the localized script. It’s now time to record, and this is where the laurel vs yanny debate takes center stage. Recording devices are now a dime a dozen. We can download dozens of recording apps for our smartphones, record directly from our laptops, set up a home recording studio, or hire a professional recording studio with all the bells and whistles. The differentiators will likely come down to available resources, time, and money, but what the laurel vs. yanny debate has brought to light (and to sound), is the need to consider how complex your message is.
Are the words that are being recorded, fairly common and easy to pronounce? Might some of the words be complex and rarely used in everyday speech? Are some of the words spelled the same in other languages but pronounced differently? You already know from reading our report on TTS that hard-to-pronounce words, or proper nouns that are pronounced differently in different languages can pose considerable problems for machine translation and require time and effort in post-production. Consider the following:
The point is, the laurel vs yanny debate really does give pause to consider just how the recipients of our message are actually hearing our message. So, take the time to truly scrutinize your script’s words.
Post-production is essential to ensure that your final product is top notch, but if laurel vs yanny taught us anything, it taught us to consider running a quick survey to determine what people actually hear when they listen to the more complicated, or rarely used words in our scripts.
Although steps 1 through 4 above are crucial to launching your product or service, you will never fully know how your message will resonate with your target audience until you deliver the goods.
If you crossed your T’s, dotted your I’s – and checked your audio localization quality to clear out the yannis that are interfering with your laurels – you should be good to go!
It isn't just for Uber drivers and AirBnB landlords. The gig economy is something that is reaching in to all other industries, and localization is no exception. In the 2017 Global Voice-Over Industry Trends and Influencers, we analyze how the rise of P2P (Pay to Play) sites for multilingual voice-overs are disrupting the established supply chain […]
In December 2020, Nimdzi was given an opportunity to test a brand new product — Spotlight. It is developed by Intento to support machine translation (MT) curation, enabling quick analysis of the MT training results. This product is intended mainly for those who train custom MT models and thus regularly face the task of evaluating MT quality.
What is language technology? Nimdzi has organized a series of panel discussions to cover some of our favorite topics in the space.
New challenges brought about by doing business in our digital world demand new solutions. Some constants still remain, however, without which a text and the quality of its translation would be less than satisfactory. One good example of such a constant is terminology and terminology management. Terminology management includes a number of different aspects, but […]