The present volume brings together detailed comparative work on a number of non-Pama-Nyungan languages of Northern Australia, and is the first book-length study to span this linguistically complex region, containing as it does perhaps 90% of Australias linguo-genetic diversity in an eighth of its land area. Many papers originated at a workshop held at the 1989 Australian Linguistics Society conference at Monash University, but several have been written specially for this volume. It has been said that no language changes faster than a proto-language, and in the intervening period a great deal of new descriptive data on non-Pama-Nyungan languages has accumulated, as well as careful sifting of complex data, which has led many of the authors to completely revise or develop their arguments since the original workshop. Hence, the delay in the appearance of the volume reflects some major shifts in position on the part of some authors. The introduction the main issues in comparative non-Pama-Nyungan studies, and forms a state-of-the-art survey of the classification of non-Pama-Nyungan languages, which have undergone substantial changes over recent decades. It also consider the main issues in their subgrouping, and their relation to the Pama-Nyungan languages. The second to fourth sections then looks at issues of subgrouping, reconstruction and areal influence that pertain to particular non-Pama-Nyungan families or subregions. The final sections returns to the issue of whether one can carry the process of reconstruction back to deeper levels than the families themselves, that is back to some level from which all or most non-Pama-Nyungan families are descended. Overall, the volume illustrates that - despite recent claims by some authors - the comparative method can be successfully applied to Australian languages. It also furnishes a number of detailed and intricate studies of morphological reconstruction applied to complex paradigms. 2003.