What was the main question in your dissertation?
Words with a different meaning – like the Dutch words ‘buur’ (neighbour) and ‘boer’ (farmer) – can differ in only one sound. In addition, there is a lot of variability in how speakers produce sounds. Even the same speaker cannot pronounce the same sound exactly the same twice. To recognise words and to understand what someone is saying, we need to perceive this variability and distinguish important differences (e.g. ‘buur’ vs. ‘boer’) from irrelevant variation (e.g. differences in pronunciation of the same word). Luckily, our brain is incredibly fast at doing so. I investigated how this is possible.
Can you explain the (theoretical) background a bit more?
Our brain needs to know which differences between sounds are important to notice. An important contrast in Dutch is the one between back vowels (e.g. /u/ as in ‘boer’, ‘farmer’) and front vowels (e.g. /y/ as in ‘buur’, ‘neighbour’). This difference is marked by a contrast in tongue position. If you try to produce these sounds, you will notice that the back or the front of your tongue is raised for ‘oe’ or ‘uu’ respectively. Another meaningful contrast is the one between rounded (e.g., /o/ in boot, ‘boat’) and not rounded vowels (e.g., /e/ in been, ‘leg’), named after the rounding of the lips.
To distinguish front from back and round from not round, listeners seem to compare the words they hear to words and sounds stored in their brain. However, previous research has shown that some features of words, for instance the feature ‘front’, are not stored. Therefore, a change from a back vowel (roos, rose) to a front vowel (reus, giant) is more noticeable to listeners than a change the other way around. This is how it works. The change from an expected back vowel to a front vowel creates a mismatch: hearing a front vowel clashes with the stored characteristic ‘back’. The reverse change, hearing a change from front to back, does not create a mismatch – because the feature ‘front’ is not stored – and the change is therefore less salient.
Why is it important to answer this question?
Understanding how differences are perceived is crucial to understanding how speech perception works. For that reason, what is stored in mental representations, how these representations are used to understand speech, and how we are able to use them so incredibly quickly are important questions in phonology. Also, this knowledge can be used to investigate whether people with a language disorder such as dyslexia or a developmental language disorder differ in how they perceive contrasts. This knowledge could be used in interventions or early diagnosis.
Can you tell us about one particular project?
In one experiment, participants heard a stream of individual sounds that was sometimes interrupted by another sound. For example, they heard an /o/ over and over again and then heard /e/. I measured brain activity to investigate how the brain reacts when the listener expects an /o/, but hears an /e/. The characteristics of the deviating vowel are compared with the characteristics of the expected vowel. This gives information about whether a feature is stored in the mental representation.
What was your most interesting finding?
As has been reported for German and other languages, Dutch listeners showed stronger brain responses to changes from a back vowel to a front vowel than to changes in the opposite direction. This strengthens the assumption that the feature ‘front’ is not stored in their brain.
More surprisingly, I found that when a sound that is not round changes into a round sound (for example /e/ changes into /ø/), this results in a large brain response, which means this difference is very prominent to Dutch listeners. The other way around, the brain did not give such a large response. This pattern differs from earlier research with German speakers.
What are the implications of this finding? How does this push science or society forward?
The divergent pattern of results for Dutch compared to German participants shows that listeners with different native languages can differ in how they perceive similar vowel contrasts.
Apparently, it does not only matter which vowels a language has, but also how exactly they are used. Detailed perception experiments that compare different languages can provide more insight into the similarities and differences between languages and language systems.
What do you want to do next?
I am currently doing applied research at Royal Dutch Kentalis. Our goal is to improve care, education and participation of deaf and hard-of-hearing children. I enjoy using my research skills in an applied setting to address issues that need a solution, and I hope this way I can contribute to make the world a better place.