This book investigates in detail the grammar of polysynthetic languages--those with very complex verbal morphology. Baker argues that polysynthesis is more than an accidental collection of morphological processes; rather, it is a systematic way of representing predicate-argument relationships that is parallel to but distinct from the system used in languages like English. Having repercussions for many areas of syntax and related aspects of morphology and semantics, this argument results in a comprehensive picture of the grammar of polysynthetic languages. Baker draws on examples from Mohawk and certain languages of the American Southwest, Mesoamerica, Australia, and Siberia.
A classic on the nature of polysynthetic languages. Using extensive data from Mohawk and other languages, Baker proposes a macroparametic view of polysynthesis, where a cluster of properties (adjunct behavior of NPs, lack of infinitives, pronoun drop, noun incorporation, discontinuous constituency, etc. etc) are all tied to the requirement that theta roles in these languages must be expressed on items (either agreement or incorporated nominals) that are part of the verb head.
The book is rich in data and classical argumentation, and provide a comprehensive description of Mohawk syntax at the same time. I was a bit puzzled at the end of the book. After nearly 500 pages of careful scientific empirical research, Baker concludes that the only possible explanation for the existence of a macro parameter is the tower of Babel. So *bam* a fine piece of linguistic research is marred by 3 pages of theist mumbo jumbo. Very disappointing.