PhD Project 11

Perception of multidimensional sounds in humans and birds: Are speech categories special?

 

PhD-candidate: Merel Burgering
PIs: Jean Vroomen (WP1) and Carel ten Cate (WP5)
Start date: 01 February 2015

(last update 2019-06-27)

Research Content

Humans are remarkably skilled at extracting various types of information from the speech signal, such as phonemic categories and speaker identity. Unclear is whether these dimensions are represented in different higher-order modules; whether and how these modules affect basic aspects of categorization; and whether these properties are unique to human speech perception or represent a more general property also present in non-human animals, which was utilised for the evolution of human speech. In this interdisciplinary project these questions are addressed by comparing humans with songbirds using similar tasks and stimuli, combining psychophysics with fMRI.

Highlight

One major result is the demonstration of vowel - and especially voice gender recalibration. Researchers demonstrated recalibration for consonants before; using a within-subject design, this study demonstrated this for vowels and for voice gender as well.

Working interdisciplinary remains a major challenge in science. This PhD project is highly interdisciplinary combining behavioural biology, cognitive neuropsychology and auditory neuroimaging. Work was performed at three different universities. The project benefitted from presentations and discussions in different research groups from different universities, resulting in new questions and fruitful feedback on the research topic.

The average proportion of FEM responses (pooled across beek and beuk) after exposure to Agender?Vmale or Agender?Vfemale (panel A) and after exposure to AVmale or AVfemale (panel B). On the X-axis, the three test sounds (A-1 (male-like), Agender?, A+1 (female-like) of the male-female continuum) are displayed. Panel C shows the overall aftereffects for ambiguous adapter sounds (recalibration) and clear adapter sounds (selective adaptation), which were significantly different from 0 (p < .01)

Progress 2018

All data collection was finished in September 2018. In Tilburg, behavioural data was collected from 15 students for a pilot (presented on poster at IMRF 2018) and afterwards 35 students (described in QJEP manuscript). In Maastricht, fMRI and behavioural data was collected from 12 participants. Currently, work is done on writing down the final analysis.

Groundbreaking characteristics

This project is very interdisciplinary combining psychophysics, biology and neuroscience. Auditory perceptual tests with humans will be performed in Tilburg, perceptual studies on zebra finches in Leiden and fMRI on humans in Maastricht.
The project is unique in combining comparative and neuroimaging approaches to study speech perception.
This study is the first to test animals in a multidimensional task and to compare the performance of humans and animals on the same stimuli. This study will contribute to knowledge about neural correlates of speech perception and cognitive mechanisms underlying speech perception/auditory perception.