Although music has always been referred to as the „language of emotions", there is an apparent lack of systematic research on the communication of emotions through music. A parallel is drawn to the established area of research on communication of emotion in the voice and, after a review of this area, it is pointed out that the questions and results relevant to this area can be transfered to the study of emotional aspects of music. As in speech research, three main questions may be asked in the study of the communication of emotion through music: 1) How is an emotional state expressed through music ( =encoding); 2) How is an emotional state inferred from music ( =decoding); and 3) If listeners can reliably identify emotions from musical patterns, what are the acoustic and formal cues used in the inference process ( = identification of cues). Relevant studies are reviewed and it is pointed out that it may be of particular interest to not only ask which formal and acoustic patterns are responsible for emotion attribution, but also for emotion induction. The empirical establishment of specific acoustic-formal profiles in music responsible for the induction of different emotional states could lead to a wide range of applications in experimental psychology, industrial psychology, marketing, music therapy and music theory. Same concrete guidelines for the advancement of this area of research are outlined.

Research on melodic construction in different categories of songs has led to the conclusion that in contrast to other song categories lullabies are not a homogeneous group. They seemed to be composed of several subdivisions. A large one, mainly consisting of bell-shaped and falling melodic contours, showed much similarity to praise songs for grown-ups, another one reminded of the rocking of a cradle. A third group with lively moving sinusoidal contours seemed to be of an entertaining character, and some lullabies actually showed features similar to those of fighting songs. The impressions while hearing these songs corresponded to the visual observations. So, for this investigation it was hypothesized that the concept of lullaby comprises songs with very different emotional contents, which are correlated to different compositions of melodic contours. Two tests were carried out to clarify this problem. The results of both experiments impressively confirm the hypothesis formulated at the start.

Expression in music in Handel's opera arias is described by the „ Affektenlehre" of the baroque period. The question is if today's subjects judge these arias in the same way as described. A questionnaire was composed of both the terms of the „ Affektenlehre" as well as of the scales used in contemporary psychology of emotion. A traditional classification was found by factor analysis and cluster analysis. Tue experimental design is the same as the one used in a former investigation (Jahrbuch d. Dt. Ges. für Musikpsychologie 1995). The specific problem which is investigated here is the effect of musical educational background.

Improvisations as a form of the spontaneous musical expression of intrapsychological processes offer accesses to otherwise difficultly reachable depth dimensions of human experience because of their high authenticity. lt was investigated whether the emotional experience of laymen will be concisely put in improvisations on the xylophone and so made recordable for listeners. 20 grammar school pupils expressed three basis feelings (anger, sadness, joy) in improvisations on the xylophone; later 74 subjects had to assign correctly the improvisations to these feelings. They managed it outstandingly often. This study has proved the intersubjectively recognizable experience of feelings or images related to feelings in improvisations of adults without musical professionality. This is relevant to the psychology of expression and has also implications for the diagnostic in clinical psychology and psychotherapy especially considering procedures in music therapy.

The subject of this study is the question whether five-year-old children are able to recognise musical motifs in an altered context. To investigate this ability to abstract, 45 five-year-old children were given five tasks in which, respectively, three musical examples were to be compared. The participants were asked which of two musical variations was most like the model sample, which always remained constant. Among the musical parameters were rhythm, melody, tone, pitch, and tempo. Contrary to the current view of musical development, the measures indicate that five-year-olds already possess the ability to abstract musical motifs. This competence correlates with the type of musical variations and gender. Finally, typical patterns of reception seem to be developed in dependence on musical socialisation conditions.sample, which always remained constant. Among the musical parameters were rhythm, melody, tone, pitch, and tempo. Contrary to the current view of musical development, the measures indicate that five-year-olds already possess the ability to abstract musical motifs. This competence correlates with the type of musical variations and gender. Finally, typical patterns of reception seem to be developed in dependence on musical socialisation conditions.

NDR 3, a broadcast wave under public law focused on classical music, got a commercial rival with Klassik Radio in 1990. In the competition for the listener's favour both stations follow different conceptions. So the quota of spoken word in the programme of Klassik Radio is distinctly lower than at NDR 3. An analysis by questionnaire of 133 regular listeners of these stations showed that besides persons listening mainly to one of them there is a large group participating in both equally. In addition to the preference for music from the classic, romantic, and baroque periods, there is a considerable demand for some background information. Moreover from the evaluation of 15 statements about listening to classical music, some factors could be extracted that can be interpreted as certain modes of music perception. These show a highly significant canonical correlation with the programme preferences. Listeners favouring NDR 3 differ from Klassik Radio listeners in preferring word contributions, correlated with a rather cognitive mode of music perception .

The purpose of this paper is twofold. First, it briefly characterizes how Piaget's theory has been mainly applied in music psychology and education to date. After discussing the maj or shortcomings of this approach, the core elements of the theory as applied to music are re-evaluated. Second, a new theoretical approach is proposed and elaborated in more detail with respect to the development of singing. The focus on singing is motivated by the fact that it is an elementary and universal means of expressing the basic musical parameters, namely the timing of pitches, from early on in ontogeny. Finally, the delineated approach is exemplified with an excerpt from the initial phase of song acquisition of two pre-school children.