英语论文网

in different contexts of musical knowledge. Implementing the musical plans in this way how little musical knowledge is actually required for surprising competence in the domain. PLANC shows how the three layers of knowledge are adequate to yield competent, yet flexible musical behaviour. MC framework is generally designed to allow novices to begin exploring interesting, motivating tasks as soon as possible.

6.1.5 The limitations of current MC implementation

Several plans, some styles, numerous pieces, the planner and a microworld for harmony have been used. The student model (a semantic net to allow construction of explanations and teaching modules) has not yet been implemented. The MC is not a practical system. It is a prototype used to demonstrate all of the key principles involved. However, the framework is compatible with standard teaching and modelling techniques developed by Baker (1994) and Cook (1994) mentioned later in this paper.

The current version of MC framework focuses on harmony, with some attention to details of metre and bass lines. However, the design of the program is quite broad and domain neutral, and can be applied to other areas of music. Levitt (1985) suggests the basis of extending the work into melody. Watson (1990) and Kane (1991) have carried out preliminary investigations into rhythm based work. The following section is based upon work with a similar framework but focuses on melody.

6.2 A constraint based learning tool for exploring melody

A constraint-based learning tool MOTIVE (Smith and Holland, 1994) explores melody in a similar context as MC frameworks, with a focus on Narmour’s (1989) cognitive theory of melody. The goal of MOTIVE is to allow ab initio beginners to explore the composition of melody. The work achieves potentially very general applicability to melody, irrespective to genre, by virtue of being based on the most fundamental psychologically grounded of theory of melody currently available (Narmour, 1989).

Narmor’s theory has very little competition as a theory of melody based upon psychological terms, however it does have its problems and limitations (Cumming, 1992). It can predict how a listener will break a melody into groups of contiguous notes and which ones will perceived as more important than others by using simple extensions to low level processing theory for melodic notes. This gives rise to hierarchical parse trees which recursively reduce the melody to simpler versions, roughly similar to Lerdahl and Jackendoff (1983) TSR trees. In order to make use of Narmour’s theory computationally, the model was refined and implemented by Smith’s work.

This enabled Smith to test Narmour’s analyses for consistency against the computational version. The test results showed the theory to be internally incoherent, with some gaps but no fatal internal flaws. This computational model then became the main component of MOTIVE. It uses a constraint based planner and is able to navigate tree plans to re-plan or recompose melodies.

Thus as well as other teaching strategies, the ‘analysis by recomposition’ strategy introduced in the MC system can be applied. The status of Narmour’s theory is unclear and whether Narmour’s theory of melody can be appli本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。