Musical experience and musical production are analysed from the point of view of cognitive self-organization theory. In the first part basic theories of perception are discussed with respect to information processing. While other theories postulate the pick-up of information from the environment or the processing of external information, the theory of cognitive selforganization assumes the generation of information in the cognitive system itself. Musical communication is consequently interpreted as a constructive process of melody, harmony, rhythm and dynamics and the attribution of emotional values within the receiver. This is demonstrated by a number of musical examples using gestalt theoretical principles of composition. lt is especially argued that music as a cognitive construction exists only within a certain time-scale window. Furthermore there are examples of multistability, showing the internal activity of the cognitive system. The principles of music generation are mainly the classical gestalt laws as there are vicinity, similarity, symmetry, closure, common motion and continuity. The constructive nature of musical experience is finally demonstrated by physically impossible sound structures analogous to Escher's examples in the visual arts. The possibility of spontaneous creation of music in collective improvisation is interpreted as a collective swinging-in by rhythmic resonance. In the last part it is dicussed how far the social context of musical production and reception defines the range of possible musical interpretations.

This study is part of a body of work devoted to the psychological organization of listening to music, certain parts of which have already been the subject of recent papers and publications (I. Deliege: 1987,1989, 1990, 1991; I. Deliege et al: 1989, 1990). The corpus itself is grounded on what is commonly called a model of the itinerary of musical information. This model - described elsewhere - is essentially based on the formation of groupings generated by cues and their imprints. In this study, I will attempt in particular to develop the idea of the cue as intemal or extemal reference to the musical work. I will then describe the aspects of the experiment itself.

The subject of the study is listeners' abstraction of themes from melodic variations. A set of transformations of a prototypical melody was generated by application of transformational rules. The prototype (theme of Beethoven's Variations WoO 65 D-major) represented the structural central tendency of the set. The participants (311 students aged 11 to 12 and 14 to 15 and a control group of college students) were given a set of transformations and a false recognition task in which they were asked to recognize the prototype and other transformations, none of which were in the original set. The measures indicated that participants abstracted a prototype while listening to the original set of transformations. A relation between the number of transformational rule applied and the saliency of the prototypes (Welker 1982) could not be established.

The purpose of the study was to determine the nature of the difference between absolute pitch and relative pitch: Is the difference one of degree, or do absolute pitch and relative pitch represent two autonomous traits? Ten absolute pitch possessors and a cross section of 120 music students were administered a prerecorded listening test. The task was to assign chromatic pitch names to individual musical pitches. The pitches in the test were of different duration (1 second, 3 seconds), performed on different instruments (piano, French horn, oboe), and there was either 2 seconds or 5 seconds intermission between two sequent pitches. Musical excerpts were interspersed for tonal reference. The listening test also contained two-tone motives, which had to be identified and remembered after 3 minutes of interference tasks, and it contained pitches whose octave position had to be determined. The absolute pitch possessors correctly named 85 percent of the pitches of the listening test, while the relative pitch possessors correctly named 10 percent - scantly above chance level. Absolute pitch possessors performed better when response time was shorter, relative pitch possessors performed better when response time was longer. Relative pitch subjects made judgments in reference to the context tonality, absolute pitch possessors did not. Differences in the cognitive strategies of absolute and relative pitch possessors are concluded.

The study was conducted to monitor the possible progress of first year and advanced music education majors in eartraining. A further aspect was to investigate the effect that computerized eartraining has on the hearing skills. Two groups of freshmen (low motivation; n=34) and one group of graduate students who were heading towards their final exams (high motivation; n=11), all from the »Hochschule für Musik und Theater Hannover«, took part in a pretest posttest experiment. One experimental group of freshmen (n=18) as well as the graduate students volunteered to work with a drill-and-practice computer software package (AudiMax) for one semester to improve their eartraining skills. The control group attended the regular theory classes and received no additional computer-based training. A specially designed listening test (TELG) that comprises five subtests - Tonality (T), Melody (M), Harmony (H), Rhythm (R), and Musical Memory (G) - was administered at the beginning of the semester and again five months later. The results indicate a significant increase in mean scores over four of the five parts of the test (F=18.92; df=4; p=.012). Subtest G was not signifikant, showing that the subjects could not remember the first testing session, therefore improvements must be attributed to regular classes and private practice. The highest gains were on subtest R, suggesting a pre college deficit in instrumental teaching. Correlations between the five subtests were significant among the H, M, and T scores, while the R and G scores neither correlated with each other nor with the H-M-T block. The results were independent of the principal instrument except for a slight tendency in favor of the pianists. The experimental group did not score significantly higher than did the control group. The high motivation group and the subjects who had scored comparatively low the first time showed a non significant tendency towards greater mean gains than the low motivation group. An interview with the students indicated that motivational factors played an important role in their attribution of score increase. Those students who had practiced in addition to their regular classes attributed their gains to these efforts. Others attributed their gains or lasses externally (e. g. luck, test). Problems with handling the computer can serve as an excuse for not working at all. Female students were as interested as the males were in working with the computer.

We could show that there is a correlation between the ability of estimating an interval of time (1 and 2 minutes) and the ability of reading a piece of music several times with a minimum of deviation of speed. This correlation concerns known and unknown pieces. The estimation of time does not seem to take the chief part in supplying the musician with a good imagination of music concerning the aspect of time (low rate of explained variance).

The study on hand was to investigate the role of authoritarian attitudes in the perception of music. Adorno's assumption that authoritarian attitudes contribute considerably to the tendency to reject modern, innovative works of art was studied in a group of laymen as weil as a group of musicians. The test material was comprised of 12 art songs (Liedern) representing various styles from the 19th century to the present day. In order to determine the music preferences a semantic differential was used which had been developed especially for this study. The findings gained from statistical procedures lead to the following conclusions: Highly authoritarian persons tend to judge negatively those musical examples towards which they have no expectations, and attribute to them a high degree of complexity and excitement. On the other hand songs which come up to their expectations are judged more positively. In strict black-and-white terms this means that the unfamiliar tends to be judged negatively whereas the familiar is judged positively.