Everyday listening and auditory icons

Abstract

This thesis represents an attempt to understand how people listen to the world, and how such an understanding can help in developing auditory interfaces for computers. Part I of this thesis is concerned with everyday listening, the act of gaining information about events in the world by listening to the sounds they make. Chapter 1 introduces this area of research and discusses its relations with other work on audition. In Chapter 2, I suggest that the differences between cognitive and ecological perspectives on perception may be understood as stemming from contrasting views on systems, and that although these perspectives are fundamentally incompatible they may both prove valuable in understanding perception. In Chapter 3, I examine what we hear, and introduce a framework for understanding the attributes of everyday listening. Finally, in Chapter 4 I investigate the perception of a simple sonic event--struck bars of wood and metal--with the aim of discovering what information for material and length is inherent in the sounds. Part II shows how the ideas developed in the first part may be applied to the use of sound in computer interfaces. The basic idea is that auditory icons, caricatures of everyday sounds, may be used to present information to users in a way that is analogous to the use of visual icons. This half of my thesis is made up of three papers that discuss my work on auditory icons. First is the paper that introduced this idea (Gaver, 1986), and which lay the framework for further research in this area. Second is a technical report written recently for Apple Computer, Inc. (Gaver, 1988), which is a survey of current techniques for using sound, and which explores the issues involved in creating auditory icons in more detail. Finally, I include a paper which describes the SonicFinder$\sp{\rm TM}$, a prototype auditory interface that I developed at Apple Computer, Inc. (Gaver, in press). This interface illustrates the use of auditory icons in an actual system.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 108 citations exist upstream.