Constraint-ranked derivation : A serial approach to optimization
H. Andrew Black · OLAC - Open Language Archives Community · 2006
A significant portion of the world's languages have the basic word order of Verb-Subject-Object (VSO).Languages with this word order and its mirror image, the rarely found Object-Subject-Verb (OSV), have long been considered problematic within X-Bar theory, since they do not easily fit into a phrase structure with a subject and a verb phrase consisting of the verb and its complements.Do these languages simply have a flat structure with no real distinction between the subject and other arguments of the verb?If so, why would languages with these word orders be so different from languages with the other four word orders?These questions became more important as linguistic theories moved to a notion of all languages deriving from a single Universal Grammar.Various hypotheses for how VSO order is derived have been proposed over the years, with the most promising being movement of the verb out of the verb phrase to a position above the subject (Emonds 1980).But exactly where are these positions?Do all VSO languages have the same analysis?Within generative grammar, the details have been further complicated by the splitting of the functional projections into multiple projections (Pollock 1989) and the idea that all projections are head-initial, specifier-initial and complement-final (Kayne 1994).This debate is the background for the book, which claims to be the first full-book study written in English of Welsh syntax using minimalist theory.(The title of the book specifies Principles and Parameters theory due to the emphasis on parameters, but the book also addresses more current issues within minimalism and the author seems to simply consider them parts of the same overall theory.)The central question being addressed is "Which values of which parameters of Universal Grammar determine VSO word order?"Various parameters have been suggested, but the author desires to limit parameters to those meeting two main criteria: they must be both typologizable, accounting for the patterns of language variation observed, and learnable, where a simple aspect of the data can be seen as a trigger for the language learner to experience the value of the parameter.These issues make the book of interest to typologists and language-acquisition specialists.A significant amount of Welsh data is provided throughout the book, making it also of interest from a descriptive standpoint.