How do we understand language?

How does our brain turn speech into meanings? Does the answer lie within learning nursery rhymes...?
09 May 2017

Interview with 

Usha Goswami, University of Cambridge


How does your brain turn speech, which after all is just a series of vibrations at different pitches and volumes, into meanings? It’s the primary method of communication in every culture in the world – but we don’t totally understand how it works. Georgia Mills went to see one group who are investigating how patterns of syllable use come into it, by putting hats on babies and playing them nursery rhymes…

Usha - I’m Usha Goswami and I’m Director of the Centre for Neuroscience in Education at the University of Cambridge.

We know that there are energy patterns in the speech signal which, of course, we’re not consciously aware of hearing but which provide acoustic landmarks for the brain so that the sets of neurons in the brain, which are basically firing electrical signals at different rhythmic rates, can synchronise themselves with those rhythmic rates in the speech signal.

Georgia - Does that mean that say when I’m talking neuron cells in your brain are lighting up in the same pattern of speech that I’m actually using?

Usha - Exactly so. In a very complicated pattern, because there are lots of different speeds of rhythm in what you’re saying and the brain is tracking all of those speeds at the same time, in a kind of hierarchy of rhythms, which then get bound together and you just perceive speech. You are not aware of all that hard work your brain’s doing in the background.

Georgia - How did we find out about this pattern of firing in the brain and why do we think it’s important in understanding speech?

Usha - This is actually very recent work, mainly with adults. I think we’re the first project in the world to look at the same mechanisms in the infant brain for babies. But it became clear without these acoustic landmarks, which are these energy changes, the brain didn’t know how to synchronise with the signal. So it didn’t know where the syllables were for example. If you take these acoustic landmarks out, then you can’t understand speech at all. If you put back just little click noises at the right point in the signal, then speech becomes intelligible again. So all of us will stress some syllables more than others; so words like mother, father, they have a strong first syllable and even if you’re whispering them there will be a stress signature or energy difference. It’s those energy patterns that the brain’s interested in.

Georgia - If you say mothER - that already sounds really weird.

Usha - And that’s something a child with dyslexia finds it hard to hear. They don't really hear those mispronunciations of stress. But we’re interested in the very beginning, the very get-go of how infant brains process those patterns of stress and unstressed syllables.

Georgia - Are these stress syllables common across all languages?

Usha - Well, one finding motivating our project is that on average all human languages produce two stressed syllables a second. So that’s a kind of acoustic skeleton that the brain can use, whichever language you’re born into, to begin requiring the hierarchical rhythms that we were talking about that were in the signal when we speak. The reason it’s the same across languages is because we all make speech the same way. We all have similar throats and larynxes, and so it’s the way our mouths operate that determine this twice a second beat.

Georgia - So how are you looking into this exactly?

Usah - We’re hoping that mothers will bring their babies in from the age of two months so that we can listen to their brain, listen to these electrical signals taking place in response to being sung nursery rhymes...

Eating her curds and whey
Along came a spider and sat down beside her
And frightened Miss Muffet away.

To hearing somebody saying a syllable at this twice a second rate… dat, dat, dat, dat. And we’ve also got a drum beat at that rate as a very sort of pure stimulus of this acoustic landmark theory.

Kirsten - My names Kirsten.

Georgia - And who’s this?

Kirsten - This is Lexi.

Georgia - How old is Lexi?

Kirsten - She is nine and a half weeks.

Usha - We collect the data by putting a sort of swimming cap of sensors onto the infant’s head and these are very like little sponges - very sensitive to any electrical activity that’s happening underneath the scalp. The computer just measures that and measures whether that synchronises itself with the different nursery rhymes, or the different drum beats, or whatever we’re playing.

Georgia - So you can look at the electrical patterns from the brain and see if they’re synchronised on the screen with what you’ve been playing out to the babies?

Usha - You can see whether they’re synchronised with millisecond accuracy. It’s a really, really temporally accurate measurement.

Georgia - Over the next few years, Usha and the team will be following up with children like Lexi to track how their language skills develop and to reveal if there is a link between how well their brain synchronised with rhythms and their growing language ability…

Usha - I think individual differences in this ability to align your brain rhythms with speech rhythms could be really important for how quickly and efficiently you acquire language.

Georgia - It seems incredible this sort of fundamental pattern in the brain that’s across everyone. Can we use this somehow to learn language faster, or help our memory, or may learn different languages?

Usha - We can definitely use it to help children who have language difficulties because different projects in my group are looking at helping children with dyslexia through learning drumming, for example, and drumming in time with speech. And also children who have speech and language difficulties can be helped by having a background pattern of beats that helps them with the phrasing of speech.

Georgia - So could the importance of synchronising to a rhythm explain why we all love telling babies nursery rhymes?

Usha - I think there’s something fundamental about metrical patterns, so poetry exploits the same thing, or Shakespeare. This patterning is very perfectly aligned when we’re stressing our syllables at regular intervals and that’s what we do in nursery rhymes with symmetrical poems…

Ride a cock horse to Banberry Cross
To see a fine lady ride upon a white horse
With rings on her fingers and bells on her toes
She shall have music wherever she goes.



Add a comment