No Blueprint, No Problem: Redefining the Speech Therapy Experience
Music has been used very effectively for children that have a stammer. Images courtesy of Rothco.
What do “Don’t Start Now” by Dua Lipa, “Good As Hell” by Lizzo and “Right Here, Right Now” by Fatboy Slim have in common? They are all being used as part of a new initiative by Rothco, Apple, and Warner Music to create ‘Saylists’, a collection of playlists aimed at augmenting the speech therapy experience for young people. (Approximately 1 in 10 children experience a speech sound disorder.)
In the field of speech-language therapy, one of the most successful strategies is the repetition of difficult syllables, words, and phrases. All 173 songs on the playlists center around commonly-challenging speech sounds: ‘CH’, ‘D’, ‘F’, ‘G’, ‘K’, ‘L’, ‘R’, ‘S’, ‘Z’, and ‘T’. “Don’t Start Now” is for the ‘D’ sound, “Good as Hell” for ‘G’, and “Right Here, Right Now” for ‘R’.
Saylists began as a passion project for one of Rothco’s creative directors, inspired by his sister, a teacher who noted how children in her class struggled with the material for working on problem sounds because they found it boring. It took 2.5 years from idea to completion: Songs were analyzed for patterns of repetition used to train specific speech sounds. An algorithm was then built to analyze the 70 million tracks on Apple Music and determine which songs fit the patterns best. Though newly-launched, Saylists are already making an impact—the University of Cambridge in the UK is exploring how they might be used in their curriculum.
We sat down with Alan Kelly, Chief Creative Officer of Rothco, and speech and language therapist Anna Biavati-Smith, to learn how they collaborated on this project that can help redefine the long, often painstaking journey that young people with atypical speech can experience.
Rothco designed an algorithm to analyze 70 million tracks in Apple Music to isolate songs with significant patterns of repetition.
What was the starting point for Saylists?
Alan: At Rothco, we run a thing called Spotlight each year. It’s a two-week period where we give our people the chance to do whatever they’d like that is outside their regular work. The idea is to put designers, creative strategists, producers, and tech people on teams to define a problem (week one) and then come up with a solution (week two).
In the Spotlight we did two and a half years ago, one of our creative directors, Rob Maguire, told us a story about his sister, a speech pathologist, who used books to help her students repeat words. The problem was the books were just overused, and the kids attention span had disappeared. Rob’s sister made up stories in her head just to keep them amused.
This was the problem Rob wanted to solve, and the team started thinking about using music and singing as a way to tackle speech issues. The more that group dug into a solution, the more they wondered, What if? If repetition can help and music can help, what if we put those two together? And what if we have an algorithm find all the songs with the appropriate speech patterns and put them together in playlists?
How does singing help with speech sound disorders?
Anna: When you sing there is a different rhythm than talking. There is a different balance between how you take that breath and how you express that word. You might say, “I like to go to the beach.” When you’re singing, the sounds are prolonged—“beeeeach.” The rhythm changes and that allows the brain to tell the mouth how to position the lips, the tongue, everything to do with articulation, and then execute. That reduces the pressure. Music has been used very effectively for children that have a stammer.
Can you walk us through “Good As Hell” and show us how you identify repetition patterns?
Anna: In this song, the sounds in repetitions are ‘F’, ‘G’, ‘B’, ‘CH’, ‘K’, and ‘T’. We have the words like, “Baby”, “feeling”, “good”, “check”, and “toss”. The most repeated word is “feeling”, which is at the end of a sentence as well as the beginning of the sentence.
And do your hair toss, check my nails
Baby, how you feelin'? (Feelin' good as hell)
Hair toss, check my nails
Baby, how you feelin'? (Feelin' good as hell) (Show it off tonight)
Hair toss, check my nails
Baby, how you feelin'? (Feelin' good as hell) (And we gon' be alright)
Hair toss, check my nails
Baby, how you feelin'? (Feelin' good as hell)
Sentences are repeated in the same way, with the same rhythm, and over time they become predictable. Predictability is an important factor in speech, especially when kids have difficulties saying the sounds. When they know that that word is in that positions and is said in a certain way, it helps them program in their mind when to say it and repeat it.
“There was no blueprint, so where do you start? You’re going in blind without a compass. You have to find your own way.”
Alan Kelly, Chief Creative Officer of Rothco
How did you two collaborate on this project?
Anna: My role was to analyze songs to see if there was enough repetition and to look at the rhythm. We don't want a song that is too fast because children with speech sound disorders substitute words rather than say the actual word. Instead of saying “rabbit” they may say “wabbit.” Or they might leave part of a syllable out of that word. If you have a word like “helicopter,” they might say “copter.” My idea was to give the children fun songs that had enough repetition in the chorus because that's the center of it.
Alan: As Anna identified the songs that would fit best, we designed the algorithm to troll 70 million tracks in Apple Music to isolate songs with significant patterns of repetition. People said, why not Spotify? The great thing about Apple Music is that they have a huge library of lyrics, so the algorithm could go in and find songs that have a good pace, have repetitive sounds, and the other things Anna was talking about. We’re fortunate that within Accenture Interactive, our parent company, we have smart scientific people who could build the algorithm quickly.
What was the biggest challenge?
Alan: A hard part was just figuring this out. There was no blueprint, so where do you start? You’re going in blind without a compass. You have to find your own way.
“Children, no matter what age, learn through movement and repetition, and music offers both.”
Anna Biavati-Smith, Speech and Language Therapist
What gave you the confidence this could work?
Anna: Music is something that everybody uses all the time. Music is a rhythm. Speaking is a rhythm. Communication is a rhythm. And emotion is related through rhythm. Children, from the moment that they are able to move inside their mom's tummy, come through rhythm. Children, no matter what age, learn through movement and repetition, and music offers both.
What response to Saylists have you seen in your work?
Anna: I gave one of the children I see 11 songs that have the “CH” sound that we’re working on. His mom told me that when he got home he went to his bedroom and explored the songs on his own time. When I saw him again, he said he had so much fun and learned the sound we had been working on. To me, that is like winning the lottery. He felt in control of what he could do and that makes a huge difference.
