Proceedings of The National Conference
On Undergraduate Research (NCUR) 2006
The University of North Carolina at Asheville
Asheville, North Carolina
April 6 – 8, 2006
Effects of Specific Formal Training on Emotional Responses to Music
Emma Lewis and Jeff Janzen
Department of Psychology
Bethel College
300 East 27th Street
North Newton, KS 67117. USA
Faculty Advisor: Dwight Krehbiel, PhD
Abstract
Are emotional responses to music altered by musical training? Though previous studies have often found no significant difference between the emotional responses of musicians and non-musicians, the effects of training may be specific to the particular piece of music studied by the participants. This specificity hypothesis was the focus of the present study. Twenty-four singers with formal training on Javier Busto’s Gloria were compared with 24 comparably trained singers from other ensembles. Participants listened to a recording of Gloria while continuously rating their emotions in a 2-dimensional emotion space (activation on the vertical axis, pleasantness on the horizontal axis). PsySound software was later used to obtain second-by-second psychoacoustic measures of the music such as loudness, dissonance, and multiplicity (number of tones simultaneously noticed). Hierarchical linear modeling (HLM) was employed to examine relationships of participants’ emotional responses with the psychoacoustic variables as well as with attributes of the participants. Results showed that dissonance and multiplicity were both significant predictors of the two emotional response variables (dissonance: negative prediction coefficient for pleasantness, positive for activation; multiplicity: positive prediction coefficient for pleasantness, negative for activation). In addition, loudness was a positive predictor of activation. Training condition significantly affected the prediction of pleasantness by multiplicity (larger effect of multiplicity in the group without specific training). While further work is needed to determine why the effects of this particular musical feature might be altered by specific training, these findings give modest support to the hypothesis that specific musical training can measurably alter emotional responses to the pieces employed in the training.
Keywords: music, emotion, musical training
1. Introduction
Multiple studies have been conducted to investigate the effects of music on emotion. In general, music has held great communicative significance throughout history, even before the time of great composers like Bach and Beethoven. From calming lullabies for infants to a warrior’s battle cry, music communicates through many aspects of life[1].
It is important to understand the shape that emotion can take when found in music. Both intrinsic and extrinsic sources of emotion can be perceived in music[2]. Intrinsic emotion comes from the heart of the music, and corresponds to structural aspects of the particular piece. It is mostly concerned with the reinforcement and violation of expectation in the listener. According to the law of good continuation, music that appears to be following a certain pattern of occurrence is likely to continue on that path. Simply stated, the idea of the intrinsic source of emotion is that when expectations are violated, negative affect is likely experienced, while reinforced expectations often result in more positive affect. However, much more detailed theories have been developed to address the many issues with such a black and white view.
Extrinsically induced emotion includes both iconic and associative sources of emotion2. Iconic sources are common ideas about expressed emotion in music, regarding a certain code of structural devices. For example, it is often believed that a faster rhythm and increase in volume imply tension, which is therefore likely to induce a feeling of emotional tension in the listener. Iconic sources are accessible by both musicians and non-musicians, and are generally agreed upon. On the other hand, associative sources of emotion are much more dependent on the personal experiences of the listener. These sources involve some connection between the music and something that is perhaps not musical at all, but that carries its own emotional attributes. The most common example is a song that reminds a person of a relationship with another, therefore inducing the emotion felt toward that person and perhaps the relationship itself. So the emotion is not necessarily coming directly from the music, but the personal associations to the music2.
A significant issue when studying emotion is forms of measurement2. The two most commonly used approaches are categorical and dimensional. Categorical measurement involves basic emotions such as happiness, anger, fear, etc., and the important task is differentiating among the completely different categories of emotion. Dimensional measurement places emotions on scales and allows for expression of amount of emotion experienced. In a two-dimensional emotion circumplex, the axes consist of pleasantness and activation, also known as valence and arousal. A primarily dimensional approach based upon this two-dimensional emotion space was used in the current study.
As illustrated, emotion colors the lives of most people to a considerable extent. One of the essential qualities of humanity, emotion is evoked by many different sources. While most people feel emotion daily, perception and expression of emotion is another matter. With regard to musical experience, emotion plays an active role; in an expressive performance, the performer both feels and expresses emotion, while the listener both perceives and feels emotion.
A major component of this study is formal training in music. Many studies have sought divergence between musicians and non-musicians, often finding exciting differences at the neural level. A study on music processing discovered significant differences between musicians and non-musicians in neural networking[3]. Participants listened to various musical sequences (e.g. random tones, and harmonized or unharmonized melodies) while undergoing an fMRI procedure. Most outstanding results showed similar activation in the superior temporal gyrus between groups, but only the musician group had significant activation in the inferior parietal lobules. Another study observed magnetic mismatch negativity (MMNm) of the melodic encoding of both musicians and non-musicians[4]. Musical training appears to enhance automatic responses to abstract changes in melodic information, while more obvious changes such as differences in tone, are automatically processed by non-musicians as well as musicians. Still more empirical work studied musical development of children, in terms of changes in the auditory cortex due to musical training[5]. Auditory evoked potentials were measured, and results showed larger P2 and Nb1 evoked responses in children receiving musical training than in those without. All of these studies combine to suggest significant neural differences resulting from musical training.
However, self-report studies on musical training have indicated less difference between musicians and non-musicians. Studies have found that humans have inborn abilities to process many materials, including music[6]. Even informal experience with music leads to development of musical competence. One study tested the ability of non-musicians to judge between well-formed and ill-formed musical phrases, finding that even without training these people can adequately process the music and recognize that which conforms to their culture’s norms. Another study examined aesthetic experiences and found no difference between musicians and non-musicians[7]. Continuous ratings of felt emotion, aesthetic response, and free response were also found to be remarkably similar between musicians and non-musicians in a separate study7. So despite clear differences in neural processing between musicians and non-musicians, studies involving self-report have not achieved similar results.
Musical training is in itself complicated. There are two distinct sides to musical performance: technicality, and emotional expressiveness[8]. Teaching emotional expression is still a topic of great debate. Can emotional expression really be taught? Some believe it is a product only of natural talent; that emotion must be felt by the performer and conveyed as the performer feels it must be conveyed. Empirical studies, however, have shown that success as a musician (while depending equally on both aspects of performance) increases with hours of practice6. Therefore, it is possible for technique to be learned and practiced.
Depending on the culture in which one was raised, music holds instinctive patterns for people. As far as the perception of emotion, most inhabitants of western culture understand a sort of expressive code made up of certain cues that performers use to express specific emotions8. The instinctive nature of this emotional code lies within the connection it has to the prosodic code, (the emotional code found within human vocal expression)6. Performers can use this emotional code to accurately express the emotion in a piece of music. The code is made up of structural cues in the music such as tempo, staccato, or loudness, that can be slightly altered to assist in emotional expression8.
There are still other ways of teaching emotional expression in musical performance8. Some teachers model the music for the student, and the student learns to play expressively through listening to the example provided by the teacher. Another method involves metaphors, or setting up a storyline for the music to induce feeling that the student must then express. These are the more traditional approaches to teaching emotional expression. Alternatives include the theoretical approach which makes use of the emotional code explained above. Cognitive feedback (CFB) is another more contemporary style of teaching. In a recent study, performers were given direct feedback about their choice of structural cues, along with explicit directions on how to more closely match their listeners’ expectations of emotional expression. The CFB resulted in fifty percent increases in accuracy of emotional expression. This method can also be employed through a recently developed computer program.
These diverse methods of teaching emotional expressiveness lend support to the notion that such expression can be taught. Gaining an understanding of the complex issues involved in musical performance aides one in grasping the nature of the present study. While many studies of emotional communication have pursued the differences between musicians and non-musicians, this study focused on two groups of musicians. If formal training can have such great effect on emotional expression, does it also affect experienced emotions? How does specific formal training on a single piece of music change the emotions a musician experiences when listening to that music? The specifically trained musicians were compared with other musicians of a similar level of mastery, who had not undergone the specific formal training. Researchers hypothesized that since training on a specific piece of music would likely increase understanding of the musicality and emotional aspects of the piece, emotional response might in some way also be enhanced.
2. Method
2.1 participants
Forty-eight undergraduate students signed informed consent to voluntarily participate in this study. They were chosen at random from the Bethel College Concert Choir, the Men's Ensemble, and the Women’s Chorus. Twenty-four were selected from the Concert Choir, who had rehearsed Busto’s Gloria for two months during daily rehearsals prior to the study. The other participants (12 from the Men’s Ensemble and 12 from the Women’s Chorus) had never rehearsed the piece before. Six were selected from each of the four choral sections from each of the singing groups. The participants consisted of 24 men and 24 women, with a mean age of 19.4 and an age range of 5 years. Mean years of private music lessons taken by each participant also involved in Concert Choir was 12, with a standard deviation of 3.9 years. Mean years of private music lessons taken by each participant not involved in Concert Choir was 3.3, with a standard deviation of 3.1 years. Mean years of participation in music groups by each participant involved in Concert Choir was 5.0, with a standard deviation of 1.7 years. Mean years of participation in music groups by each participant not involved in Concert Choir was 4.9, with a standard deviation of 2.6 years.
2.2 instruments
Participants listened to a piece of music composed by Javier Busto entitled Gloria as performed by the Bethel College Concert Choir (including 24 of the participants). A Macintosh computer running a LabVIEW virtual instrument loaded the Gloria sound file (wav format) and played it through headphones to the listener. As the participants listened, they continuously reported their emotions using a computer mouse that manipulated the LabVIEW virtual instrument. LabVIEW displayed the 2-dimensional emotion space proposed by Barrett and Russell[9]on the computer monitor. Barrett and Russell proposed that primary emotion, the subject of the current study, is entirely 2-dimensional; factor analysis has supported this proposition[10]. Thus, using the 2-dimensional emotion space was a most efficient way to gather data on what people were feeling as they listened to the music.
PsySound software[11] was used to read the digital music file and calculate second-by-second measurements of the following psychoacoustical aspects of the music segment: dissonance, multiplicity, and loudness. Dissonance is the sound of clashing, when sounds do not complement each other musically. Dissonant sounds may also be described as uncomfortable or unpleasant, because they are not easy, soothing sounds. Multiplicity is the amount of different, unique sounds occurring simultaneously in the music. Loudness is a relative volume rating of the music. A form was administered for debriefing. This standard questionnaire helped assess potential confounds, including years of private music lessons taken, years of participation in music groups, and reactive effects.
Data were analyzed in 2 levels with Hierarchical Linear Modeling (HLM). Level 1 data effects consisted of the numerical results of subjects' emotion ratings during listening as well as the PsySound ratings of the psychoacoustical aspects of the music, while the socio-demographic information made up the level 2 data file.
2.3 procedure
Each participant arrived according to a scheduled appointment, and the experimenter read the instructions to him/her. The participant then completed a practice phase during which he/she rated his/her emotions while listening to two separate pieces of music. No data were collected during the practice phase of the experiment. After this phase, the participant was asked several questions to check his/her understanding of the procedure, the meanings of important terms such as activation and pleasantness, and specifically, the difference between pleasantness and liking. Once complete understanding was established and ensured, participants proceeded with second phase of experiment during which they listened to Gloria as they continuously reported their emotions. After the completion of the song, participants filled out the debriefing form.
2.4 data analysis
The LabVIEW virtual instrument created a spreadsheet-readable file containing emotion space coordinates for each second of the music. These data were combined in a spreadsheet with the second-by-second psychoacoustical calculations of PsySound. The data from different participants were appended to create a single large data file of the emotion and psychoacoustic measurements (level 1 file). A separate file of data from participants' debriefing form was also created (level 2 file).