We begin this chapter by building up a general framework of rendering rules and styles in Section 4.1. Rendering rules enable the generation of different audio views of the same object. Section 4.2 introduces the concept of fleeting and persistent sound cues and uses these to develop an audio layout for conveying structure present in typical document content, e.g., sections and itemized lists.
We then develop a system of audio notation to convey mathematical content and present a collection of rendering rules for producing such audio notation in Section 4.3. The processing required to produce context sensitive renderings from the quasi-prefix representation (see Chapter 2.2) is covered in Section 4.4. Some descriptive rendering rules for mathematical content are discussed in Section 4.5.
When reading complex mathematics in print, we often use a chunking strategy: rather than read the entire expression, we first obtain a high-level view by grouping related subexpressions. This becomes even more necessary in oral communication —otherwise, the listener has to remember the mathematical content. Complicated mathematical expressions can be better communicated by first substituting names for complex subexpressions to provide a quick overview. This process, called rendering with variable substitution, is described in Section 4.6. A flexible mechanism for delaying the rendering of specific objects is described in Section 4.7. Appendix A.4 documents the external interface to the system of rendering rules and styles.
Throughout this chapter, we describe how AS TE R renders different object types. Note, however, that the user is free to customize all of the renderings produced by AS TE R.