Discovering history through language: Papers in honour of Malcolm Ross (review)

This volume is dedicated to Malcolm Ross, a prolific scholar who started in linguistics later in life and who has made significant contributions to Austronesian (An) and Papuan linguistics. The volume starts out with an overview of Ross’s life written by his former student (and volume editor) Bethwyn Evans, beginning with Ross’s early interest in language. His first career as an English teacher took him to Papua New Guinea, where he became interested in An and Papuan linguistics. Although he began his PhD at the age of 40, Ross has become a very influential scholar in the studies of An and Papuan languages, particularly in historical linguistics. The second chapter of the book is a complete listing of his works, including unpublished papers.

There are two main sections in this volume. The first, entitled “Historical relationships amongst languages,” contains ten chapters dealing with subgrouping and classification, mostly on An and Papuan languages. The second section, entitled “Historical development of languages across time,” contains twelve chapters, mostly dealing with reconstruction and language change, and primarily focused on An languages. Because of the number and varied content of the chapters, this review can only touch briefly on each chapter.

The first main section begins with a chapter by Roger Blench, who examines various kinds of evidence (linguistic, archaeological, zoological) for early presence of An peoples in areas where An languages are not spoken today, including Japan, China, Thailand, Australia, Africa, and the Americas. Some hypotheses he rejects, such as those linking Austronesians with the main Japanese islands, but others he considers more convincing, such as the presence of Austronesians in the Americas before European contact.

The next contribution is from Robert Blust, who brings together six different Austronesian etymologies to show how single words, or more accurately, single cognate sets, can shed light on questions of linguistic subgrouping. For example, Chamorro retains Proto-Austronesian (PAn) *baRiuS ‘typhoon’ as pakyo with the original meaning. This etymon is often retained with the meaning ‘wind’ or ‘strong wind’ in areas that do not have typhoons. The simplest explanation for the Chamorro reflex, however, is that it never lost its original meaning ‘typhoon’, implying that the ancestors of today’s Chamorro speakers never migrated through an area without typhoons. This single word points strongly, then, to the northern or central Philippines as a source area for the Chamorro language.

The next two chapters follow up on genealogical subgroupings originally proposed by Ross. Bethwyn Evans’s paper is a first attempt to show the relatedness of the South Bougainville languages using the comparative method. This small family of Papuan languages was proposed by Ross (2001, 2005) based on lexicostatistical, structural, and pronominal evidence. Evans examines published and unpublished lexical data and proposes correspondence sets and reconstructs lexical items in various semantic domains. Evans herself acknowledges that it is a first attempt, one that is hindered by gathering data from various sources and a lack of documentation, particularly of the phonologies and lexicons of these languages. However, this bottom-up reconstruction is an extremely valuable first step in testing claims of genealogical relationships.

In the next contribution, Alexandre François presents data from three An languages spoken on Vanikoro in the Solomon Islands (Lovono, Tanema, and Teanu). He shows how these languages are highly similar in structure while being quite divergent in their lexicons. Although such situations are relatively common and are usually considered to represent genetically divergent languages converging in their grammars under heavy contact, François suggests that this situation is different. He claims that the languages are closely related genetically and have diverged phonologically. The chapter, however, fails to demonstrate this close relationship and suffers from a lack of systematic sound correspondences. The data presented to show the structural similarity between the three languages are quite striking, but one is left...