1 Musical Networks: Parallel Distributed Perception and Performance Niall Griffith and Peter M. Todd, editors MIT Press/Bradford Books Preface Some 50,000 years ago, our Neanderthal ancestors may have created the first artifacts-- bone flutes--to help produce tonal music (Wilford, 1996). About a decade ago, Rumelhart and McClelland (1986) first widely introduced a new set of computational tools--parallel distributed processing (PDP) models of human cognition--that could help us further understand the musical legacy of the past 50 millennia. Five years later, the first book on the musical uses of these new computational tools was published (Todd & Loy, 1991), and now, after a similarly short period of time, we present the current volume with the hope of summarizing the latest work in this field. What is this field? The authors of the chapters in this volume, while coming from a variety of disciplines, all aim at greater understanding of the processes by which we make and listen to music. These processes are all assumed to be implemented in cognitive and perceptual mechanisms, operating on different representations of music in our brains and bodies. To learn more about the processes and representations involved in music perception, production, comprehension, and composition, the researchers gathered here make computer models based on “brain-style computation” to see if we can accurately capture human musical behaviour in an artificial system. The extent to which we succeed reflects our level of understanding of the musical phenomena involved. The models all use some form of PDP, patterned after the way that the multitudes of interconnected neurons in the brain process information. These so-called “connectionist” or neural network computer models allow investigation of processes, such as learning and generalization, and forms of representation, such as non-symbolic distributions of activity, that were difficult or impossible to study in earlier psychological models. Hence, the “musical networks” in this book are all intended to help us explore and learn more about the ways our minds--and brains--behave when perceiving and thinking musically. Most of the articles collected here come from a special issue of the journal Connection Science that we edited on the topic of music and creativity. We have added recent articles on connectionist approaches to musical behaviour that have appeared in other journals as well, and new chapters written specifically for this volume, to make our coverage of the current state of the field as complete as possible. The research presented spans a broad spectrum of musical activity, from the perception of pitch, tonality, meter, and rhythm, to memory and processing of melodic structure, to the creative processes underlying composition and harmonization. As such, we think that there is much of interest in these chapters for readers studying and working in a wide range of cognitive science fields, including perception, psychology, neuroscience, artificial intelligence, music, and