Page  00000001 Bridging the Gulf Between Music and Listener: Effect of Style Label and Descriptive Notes on Listeners' Ratings of Understanding, Liking, and Artistic Merit of Computer Music Robert J. Frank, D.M.A. & Willi Steinke, M.M. Division of Music, Meadows School of the Arts, Southern Methodist University, Dallas, TX 75275 I. INTRODUCTION "Electro-acoustic" or "computer" music is perhaps the most misunderstood and least appreciated of musical genres. Even after a half-century of existence, composers and listeners still struggle with terms to describe it and the ideal conditions in which to present it. It is well-known that in deriving meaning and emotion from music, listeners rely heavily on past experience to derive specific expectations. Many listeners have acquired a set of expectations regarding the terms "20th century music", "modern music", "electronic music" and the like. Such expectations often compare unfavorably with other terms such as "classical music", "popular music", etc, leading many listeners to be predisposed to dislike any music designated with terms like "modern music". Although empirical research in this area is limited, a small number of previous studies have found that the same music may be perceived quite differently when listeners are given different style labels, such as "classical" versus "popular" (Geigor, 1950), or bogus information about performers, composers, and prior listeners' opinions (Radocy, 1976). It has also been demonstrated that listeners who received musical as opposed to extramusical information about a classical work showed an increase in preference (Larson, 1971). It is thus possible thatjudicious choice of terms and presentation of information about the music to be encountered may enable listeners to overcome existing prejudices about music, especially unfamiliar sounding music, and possibly lead audiences to a greater acceptance, understanding, and appreciation of new artforms. The present paper seeks to further this research in the field of computer music by examining the influence of style/ genre labels and presence of title and information about the work on ratings of understanding, liking, and artistic merit. In this experiment, 16 pieces of computer music were presented to 32 participants. Each piece was presented an equal number of times under each of eight conditions: four different style labels ("computer music", "choreographed sound", "sound sculpture", and "no label"), and with and without title/descriptions for each style label. Thus, any differences in listeners' ratings of these pieces when presented under the different experimental conditions may be attributed to the effect of external factors, rather than to the intrinsic qualities of each piece itself. II. METHODS Participants Participants were 16 musicians and 16 non-musicians recruited through posted advertisements or by their classroom instructors. Fourteen were males and eighteen were females, with an average age of 21 years (range 14 -51). Twelve participants reported no previous exposure to this type of music. Participants represented a variety of geographic, ethnic and socio-economic backgrounds. Materials Sixteen pieces of music, representing a general sample of recently composed electronic music, were included in the study. Pieces were selected from the computer music literature, from a variety of composers, and represented a variety of styles (MIDI, Electronic Music, musique concrete/ acousmatic). Pieces consisted of works or movements/ excerpts that were shorter than three minutes in duration and had an average duration of two minutes (see Appendix A for a complete listing of pieces and durations). When excerpts were used, complete sections were selected. A brief description of each of the 16 pieces was prepared by the first author, based on previously published program notes where available. The sixteen pieces were recorded in two random orders onto a compact disc using SoundEditl6 and Toast software on a G3 Macintosh system. The pieces were played for participants in free-field using an RCA compact disc player, a Denon PMA-320 stereo amplifier, and B & W 500 Series loudspeakers. Procedure Participants were tested in small groups in a classroom setting. Each participant was given a brief description of the procedure and asked to sign a consent form. Participants were then given an ordered set of 16 data sheets, asked to carefully read the instructions on each sheet before listening to each piece of music, and asked to wait until each piece was over before making their ratings. Pieces were played at a comfortable loudness level, and the disc player was paused after each piece to give participants time to make ratings and read the instructions for the next piece. At the conclusion of the music listening, participants were asked to fill out a demographic sheet and post-test questionnaire.

Page  00000002 Independent Variables A four by two factorial design was employed for this study, consisting of four levels of Style Label (Computer Music, Choreographed Sound, Sound Sculpture, No Style Label), and two levels of Title/Description (title and brief description, no title or description), for a total of eight conditions in all (see Table 1). Each of the 16 pieces was played once per listening. Each piece was evenly assigned one of the eight conditions for each of the 32 participants (participants were not made aware that others in the session were listening to the same works with different style labels and/or title/description information). Randomization procedures were used to counterbalance the order of presentation of the 8 conditions within each participant, and to counterbalance conditions across pieces. III. RESULTS Title/Description Measure Understanding Computer Music Choreographed Sound Sound Sculpture No Style Label Liking Computer Music Choreographed Sound Sound Sculpture No Style Label Artistic Merit Computer Music Choreographed Sound Sound Sculpture No Style Label Yes No Condition Style Label 1 Computer Music 2 Choreographed Sound 3 Sound Sculpture 4 None 5 Computer Music 6 Choreographed Sound 7 Sound Sculpture 8 None Title/Description No No No No Yes Yes Yes Yes 3.8 (1.9) 4.2 (1.7) 3.8 (1.7) 4.1 (1.6) 3.8 (1.9) 3.9 (1.6) 3.9 (1.8) 3.9 (1.7) 4.0 (1.7) 4.2 (1.5) 4.1 (1.6) 4.3 (1.5) 3.9 (1.8) 3.3 (1.7)** 4.0 (1.8) 3.5 (1.7) 4.2 (1.8) 3.2 (1.7)** 4.3 (1.8) 4.0 (1.8) 4.2 (1.6) 3.6 (1.6) 4.5 (1.7) 4.1 (1.6) Table 1: Design of the experimental conditions (for each piece). Dependent Variables The instructions on the data sheets varied according to the eight conditions. For example, under Condition 1 participants were instructed as follows: "Please listen to and then rate the following piece of Computer Music on each of the following questions", while Condition 5 had the same instructions and added the title/description of the piece. For each piece participants were asked to rate "Your understanding of this piece" on a scale of 1 to 7, with 1 indicating "no understanding/piece made no sense", and 7 indicating "well understood/piece made good sense". Overall liking was similarly rated from "no liking" to "liked it a lot", and artistic merit was rated from "no artistic merit" to "high artistic merit" on the same 7-point scale. The order of the three ratings was counterbalanced within each subject. The post-test questionnaire asked participants to rate their understanding, liking, and artistic merit for the 16 pieces overall, and to indicate which of the terms "computer music", "choreographed sound" or "sound sculpture" best described the pieces. Note. Standard deviations are presented in brackets. ** p<.01 Table 2: Mean Ratings for Understanding, Liking, and Artistic Merit. Analyses of main effects showed no prominent effects for either Style Label or Title/Description (see Table 2). However, when the 16 pieces were labeled "Choreographed Sound" (CS) they were rated significantly lower on understanding and liking when no title/description was given than when a title/description was given. When a title/ description was given, ratings were similar for all style label conditions. In addition, when the 16 pieces were labeled CS and not provided with a title/description they were rated significantly lower (p<.05) on understanding, liking, and artistic merit in comparison to the ratings given for the same pieces when labeled "Computer Music", "Sound Sculpture", or not given any label (Conditions 1,3,4). The results did not show any consistent differences in ratings between musicians and non-musicians, however, previous exposure to electronic music did yield differences. There was a modest but significant trend for participants who had heard electronic music before to give higher ratings for understanding and artistic merit than participants who had not heard electronic music before (ratings of 4.1 vs. 3.5 for understanding (14.6% higher), and 4.3 vs. 3.8 for artistic merit (11.6% higher). The mean rating of all 32 participants for overall understanding of the 16 pieces on the post-test questionnaire

Page  00000003 was 3.9, a score that was similar to the overall mean rating of 3.95 for understanding given to individual pieces under the various conditions. However, participants gave a higher overall rating of 4.5 for liking on the post-test questionnaire than the overall mean rating of 3.9 given for liking of individual pieces. Participants also gave a higher post-test rating for artistic merit, 4.9, than the mean rating of 4.1 given for individual pieces. "Computer Music" and "Sound Sculpture" were each chosen by 15 participants as the best terms to describe the 16 pieces, while "Choreographed Sound" was chosen as the best term by only 2 participants. IV. DISCUSSION Although the overall results did not reveal consistent main effects for style label or title/description, this may have been due to conflicting scores that cancelled each other out between subjects. In order to correct this, individual scores were compared against each subject's global, "overall" ratings. The following chart presents the difference between subjects' overall rating for each category (understanding, liking, and artistic merit) minus their scores for that category. Negative numbers indicate the amount that scores were below their overall ratings, and positive numbers reflect the amount that ratings were above the overall rating. From this data, shown in Table 3, several interesting effects can be observed. without title/descriptions: U +.03 L -.23 A -.56 TOTAL -.26 0. -0.2 -0.4 o -0.6 u -0.8 II -1 -1.2.. w/o Title/Desc. I I with Title/Desc. j Figure 1: Overall effect of each condition point on average. However, the presence of a title/ description, actually increased understanding for this style label. This inverse effect seemed to influence each of the other conditions to a lesser extent. The presence of title and descriptive information seems to lessen the effects of a style label for the "liking" and "artistic merit" evaluations, as noted by the much more consistent results for those conditions. However, without program notes, the effect of a style label becomes much more prevalent. Without program notes, "Sound Sculpture" and "Computer Music" received dramatically higher ratings. This was also reflected in the follow-up questionnaire, where subjects overwhelmingly selected "Computer Music" (15) and "Sound Sculpture"(15) as the best term to describe what they had heard, versus "Choreographed Sound" (2). There results would seem to indicate that listeners understood and enjoyed works under style labels that focused attention on timbre or means of production (Sound Sculpture and Computer Music) rather than narratization or spatialization (Choreographed Sound). "Sound Sculpture" received the highest rating, and in discussions afterward, many participants also expressed the opinion that including the term "music" in the style was perhaps too limiting, carrying with it associations to the established tradition of acoustic works. However, because of the narrow difference in the results, the established term "Computer Music" would seem to also serve as an effective label for this genre. Most surprising were the results for the presence of a title and descriptive notes. The drastic difference in the above chart (Figure 1) indicates that presence of title/description does have a large effect on listeners, however, the differing effect of the various style labels tended to cancel out this effect, resulting in a more "average" overall mean. Also worth noting: one of the highest scores for "understanding" was achieved when only a title/description was given. The average effect of conditions 5-8 (presence of title/ descriptions) was -.44, while condition four (no style or title/desc.) was -.51. Also, when evaluating the raw data for individual pieces, the presence of a title/description (conditions 5-8) accounted for the highest "understanding" rating for 11 of the 16 works (piece "H" had a tie). These highest rated conditions, however, varied from piece to piece, thus averaging out with each other and reducing the main effect in the overall results. I I I I 2 -.58 -1.25 -1.19 -1.01 3 +.14 -.08 -.34 -.09 4' 4 -.38 -.47 -.69 -.51 with title/descriptions: U - L - A - TOTAL - I I I I 6 +.26 -.61 -.77 -.37 7 -.08 -.63 -.80 -.50 8 +.25 -.69 -.67 -.37 Table 3: Average effected change on each condition The greater incidence of negative scores indicates that subjects perceived their overall experience higher than their piece-by-piece ratings, especially in their liking and artistic merit evaluations. While understanding is primarily a left brain (logical) process, the more emotional, subjective ratings of liking and artistic merit increased upon reflection afterward. This would seem to indicate some degree of personal growth by the listeners, and is what one would hope would happen when reflecting upon art. The style label "Choreographed Sound" had the greatest negative effect on liking and artistic merit when used without the presence of a title/description, lowering scores by a full

Page  00000004 1 2 3 4 5 6 7 8 A B C D E F G H I J K L M N O P -3 -2 2 -5 -5 2 2 3 -3 -4 2 -5 -12 3 0 -2 -1 -4 1 -2 -5 -4 7 0 -2 -5 4 -2 2 -3 -1 5 -1 1 0 0 0 2 -4 1 2 3 0 4 2 3 -7 6 5 -2 -2 -7 -6 -5 -2 -5 5 -2 2 -5 0 5 -3 -2 -5 6 6 -4 6 9 5 3 -7 0 1 -4 -1 -2 -2 -4 -2 -5 -2 8 -1 -6 -5 -1 10 0 3 6 4 5 1 6 5 -6 -8 2 6 5 2 4 -2 -5 -3 1 1 -4 -2 0 1 -6 -1 -5 3 -1 1 0 0 -6 4 -6 2 8 3 2 Table 4: Accumulated effects on "Liking" on individual pieces for each condition (highest scores for each piece shown in bold) Individual analyses of ratings of the 16 pieces also indicated that several (pieces B, G, N, and 0) received relatively low ratings for understanding, liking, and artistic merit, whether title/description was provided or not. Similarly, several pieces (E, F, and I) received high ratings whether title/description was provided or not. An analysis of these works consistently showed that the highest scoring works contained a higher amount of stable, definite pitch (indicated in the spectrograph as horizontal marks in figure 2) verses low scoring works which contained very little if any definite pitch materials (figure 3). V. CONCLUSIONS As this art form crosses into a new millennium, composers of computer music are faced with the challenge of presenting their works in a manner that best allows listeners to understand and appreciate them. In the present study, it is encouraging to see that listeners are rating their understanding, liking, and appreciating the artistic merit of these works at higher than average levels. Post-listening discussions and evaluations were overwhelmingly positive. Descriptive information, such as titles and program notes, does have a positive effect on listeners, although when paired with an inappropriate style label, title/description information can actually detract from the listener's experience. In the informal discussions that followed each session, most participants said that they felt that descriptive notes or titles were important in establishing a context into which to put these works. Style labels that focus upon narratives or spatial aspects, such as Choreographed Sound, resulted in significantly lower ratings unless paired with title/descriptions. However, both the terms "Sound Sculpture" and "Computer Music" would seem to have a positive effect compared to no label at all. In summary, appropriate use of style labels and piece descriptions does appear to effect the understanding, liking, and artistic appreciation of this type of music. As composers of the third millenium continue to present their works, careful thought as to how these works are described and presented will most certainly help bridge the gulf between the music and the listener and allow for the greatest understanding and appreciation of this exciting art form. REFERENCES Geiger, T. (1950). A radio test of musical taste. Public Opinion Quarterly, 14, 453-60. Larson, P. (1971). The effect of musical and extramusical information upon musical preference. Journal of Research in Music Education, 19, 350-54. Radocy, R. E. (1976). Effects of authority figure biases on changing judgments of musical events. Journal of Research in Music Education, 24, 119-128. APPENDIX A: ALPHABETICAL LISTING OF PIECES/EXCERPTS: Chris Banner: "I Can't Stand This Place"; Jeff Bell: "Clip is Good"; Robert J. Frank: "Binary Blizzard"; Robert J. Frank: "Muse"; Howard Fredricks: "Screams from Lima"; Kim Hagen: "Unzipped"; Kim Hagen: "White Noise"; Jonty Harrison: "... et ainsi de suite...: Mvt. I i propos, Mvt. II (parenthese) & Mvt. V r6flexion"; Elainie Lillios: "Sounds from the Second Floor: Mvt. I - Yesterday"; Adrian Moore: "Superstrings (2 excerpts); Michael Thompson: "Miniatures (excerpt/final movement)"; Yu-Chung Tseng: "Electronic Etude No. 1 (opening excerpt)"; D.J. Sullivan: "Soundpiece No. 1" Figure 2: Spectrograph of a representative portion of a high-ranked piece Figure 3: Spectrograph of a representative portion of a low-ranked piece