Exploratory research on speech perception in amusics

An interview with Jasmin Pfeifer

1 July 2016

Jasmin Pfeifer is a PhD-candidate at the Amsterdam Center for Language and Communication supervised by Paul Boersma and Silke Hamann. She is originally from Düsseldorf, Germany, where she completed my Bachelors and Masters degree in general linguistics. Her research is about Congenital Amusia, a neuro-developmental perception disorder.

“I did my bachelor’s and master’s in General Linguistics in Dusseldorf, where Silke (Hamann) used to work before she came here. She was my supervisor for both my Bachelor and Master, and I wanted to do my PhD with her; she moved here, so I followed. But I am still maintaining my affiliation with Dusseldorf: I spend half my time here and half my time in Dusseldorf. That is also because of my participants: I have a group of participants I test in Germany and a group of participants I test here. So had I moved here, I would have still had to commute to Germany to test the German participants. So I kept my apartment in Dusseldorf, which is about 2 hours from here, and there is a direct high-speed train. The traveling is not even so intensive, because I get a lot of work done on the train, since the train is usually very empty, and there is no internet or phone to distract me. I am here half the week and the other half in Düsseldorf.. I really like this double position; I really love getting to know the Dutch university culture. Here, everything is structured slightly differently, with research institutes, and the teaching being structured in different ways. In Germany, the teaching and the research are structured and done by the same people; it is not done by two different institutes. And because I am affiliated with two universities, I also benefit from the possibility of attending workshops and conferences at both universities.”

“I am working with Congenital Amusia, which is an auditory perception disorder that negatively affects pitch and rhythm perception. It has mostly been research for music, and I am doing research on it concerning speech perception. I am looking at how amusics’ speech perception is negatively influenced by this disorder. I already did my Bachelor’s and Master’s thesis on this. For my Master’s thesis, I looked at how intonation perception is impaired by amusia. In my PhD I am trying to find out the exact speech cues that might help or hinder the perception of amusics. I look at vowel contrasts, and how amusics can or cannot perceive vowels, and I also look at word stress. I do think that the speech perception of amusics is somehow impaired; otherwise I would not be doing this research. It is a very interesting topic, as there has not been much previous research on speech perception in amusics. There has mostly been research on music cognition, or some on emotion perception in speech, and tone perception in Mandarin amusics. But there has not been any research on word stress  perception in amusics.”

“I have been wanting to do research in linguistics ever since I started studying. When I was in primary school, I wanted to be a teacher. Then my teacher took me and stuck me in front of the class and made me teach a lesson to all of my classmates. After that, I did not want to be a teacher anymore, because I was just petrified. But as soon as I started studying, I immediately liked linguistics so much that I realized I want to stay in this field forever. After that, I never really considered doing anything else anymore.”

“Rather by chance, I also got to do a fascinating tACS (Transcranial Alternating Current Stimulation) study two years ago, where we looked at amusics’  pitch memory. We tried to influence amusics’ pitch memory with electric brain stimulation, which was a really interesting technique to learn. It has been shown that amusics may have different symptoms: pitch perception deficits, a rhythm deficits but also tonal memory deficits are possible. A previous study showed that a certain frequency band is absent in amusics. We tried to induce that frequency band artificially in their brain waves. When we did this, it actually improved their memory performance for tonal sequences.” However, the amusics really only have memory deficits for tones; we tested this through a visual control sequence. It actually took us a long time to come up with a visual control sequence that is not influenced by phonology. At first we figured we could just take numbers, because that is a common visual control task. But then people say ‘one-two-three-four-five’, and you have a phonological component. If you take pictures, like a house or  a tree etc., you have the same problem. We ended up using Devanagari, an Indian script for writing Hindi and Sanskrit. If you do not know what the Devanagari signs mean, they just look like squiggles to you. Then, you will just think ‘squiggle-squiggle-squiggle’, which will not really give you a phonological representation. We showed that for this visual span task, their memory did not improve. So we showed that this certain frequency band is really associated with amusics’ tonal memory.”

“In experiments like these, you have to make sure that the tone differences are above the perception threshold. This threshold differs from person to person. Therefore we used pre-tests to determine amusics’ and controls’ thresholds. If you take pure sine tones, amusics and controls  are actually able to perceive a 0.3- 0.4 semitone difference.. However, the direction of the change also matters and amusics are worse than controls in discriminating the direction of a change: they can say two tones are different, but they cannot say in which direction they are different for example. Or they are worse in one direction than in the other.”

“Not every amusic is impaired equally in pitch and rhythm, that would be too easy. It is a very heterogeneous disorder. In the most researched cases, people have both a pitch and a rhythm deficit. But there are also people who only have a pitch deficit, and lately there have been people reported with only a rhythm deficit. Part of my research was about the diagnosis and screening test for amusia. There is only one screening device that is used, the MBEA (Montreal Battery Evaluation of Amusia), and this is not ideal. We wrote a couple of articles about why it is not ideal, and proposed some solutions. For example, there are three pitch-based tests, and two rhythm-based tests. One of the rhythm tests is more reliable, but the second,  a meter test presupposed knowledge of the march and the waltz, which a lot of people do not have: many people in the normal population also score low on this test. So there are actually three pitch-based tests versus one rhythm-based test. This makes it difficult to find the people with a rhythmic impairment, but they are surely there. It is actually quite remarkable: these people cannot even clap along properly with a metronome, which gives a regular beat. It has been shown that humans can actually anticipate this regularity, and clap some milliseconds early, but amusics cannot. Another thing about the MBEA is that it is really a music test. As a linguist I would of course also be in favor of not using musical pieces to diagnose people, but use purely pitch-based or rhythm-based tests, that use sine tones for example.”

“There has been some neural research on amusia. A lot of those studies are a bit problematic though, because they use MRI or fMRI with only a very small number of  subjects, which is really not enough for this type of study. It was always thought that the impairment must be in the auditory cortex, because it is an auditory perception disorder. However, there was no impairment found in the auditory cortex, but there were differences in the Broca’s and Wernicke’s area. This again might point to a speech related impairment. Recent studies have shown that the impairment might be in the arcuate fasciculus, the connecting fibers between Broca’s and Wernicke’s region. Unfortunately it is the case that there is a study with 5 subjects that finds something, and then a year later another study, also with 5 subjects, finds exactly the opposite. So there is a neural background, but it is all still rather unclear. This is the same for the genetic background. It is said that amusia is hereditary, but this has never been studied formally. There is one study where people drew up family trees, and pointed out the family members that might have had amusia, but these familiy members where never tested. I actually also have some family members in my participant group, so there are definitely indications, but it is not scientifically proven yet.”

“It is said that 4% of the population has amusia, but this is a questionable number. It comes from a study in the 80s, where the ‘distorted tunes test’ was used. This test uses popular melodies, in which wrong notes are inserted. People have to say whether there was a wrong note in the melody or not. However, in this study, there was not even an example of the correct melody, so the participants had to know the songs previously. Also, they recorded everything on a tape recorder, and put 40 people at a time in a room, which is also not the best scientific experimental setting. So it was very noisy data, and they did not test what they were supposed to test. I would say it is more like 2-3% of the people, or maybe even less. This is also why it has actually taken me so long to find participants; that is actually the hardest job.”

“Amusics’ everyday communication is also unimpaired, it seems, just like with non-tone language speakers. Only when you put them in a lab and take away all other cues, like context, facial cues, etc., their speech perception is impaired. Amusics are very aware of their musical deficits – that is also how I find them, but they are not aware that they also have speech impairments. But then when I start asking questions, a lot of participants admit that they have trouble perceiving irony for example, because that is sometimes only marked by intonation. When I tell participants they have amusia, a lot of people are actually relieved. They have been struggling with it all there life, and nobody could tell them what was wrong. A lot of these people are really desperate, and then it is a huge relief that they actually have a disorder. Usually you would say that amusia does not have such an effect on your life, because it just influences music. But actually, it does really play a role. In the most extreme cases, amusics feel annoyed by music, and music is really everywhere in our daily life. In the most extreme cases it can even cause headaches.”

“There is only a handful of groups that are working in amusia: there is a group in England – or actually two by now, because they sort of split up – there is one group in Canada, who actually founded this field, which is run by Isabelle Peretz, who also coined the term amusia. There is one group in China, working with tone languages, there is one group in Australia, and one in France, and that is it. So I basically a really small field. There are no real conferences or workshops purely on amusia, people usually gather at music cognition/music neuroscience conferences. We hosted a workshop on amusia and  other disorders during the SMART 2015 conference, and we will be hosting a SMART workshop purely on amusia this October. But usually there are no specific amusia workshops, and among linguists it is hardly or not of all known.”

“In my free time I go horseback riding; I have been doing that since I was about 6 years old. I spend my weekends horse riding, and I do that 2-3 times during the week as well. I used to train horses and give riding lessons; I actually spend 1,5 years in Canada doing just that. And I love traveling, which is also why I like this job so much, because I get to travel around. I also like photography, and rock climbing.”

by Gisela Govaart, June 2016

Published by  SMART Cognitive Science