As long as I have studied languages, I have known that most languages do not possess a direct equivalent of the verb "to have"; however, I learned very little about how this majority of languages actually expresses the concept of predicative possession except to see that they would express it using a simple locational adverb. In the brilliant World Atlas of Language Structures, however, it is shown that there are two further types of predicative possession that will not be familiar to speakers of the world's major languages: the "topic possessive" and the "with possessive". Although the WALS database shows some clear tendencies in terms of linguistic typology: locational possessives are found in case-marking languages, topic possessives in isolating languages and "with" possessives in polysynthetic head-marking languages, it does not do anything to explain these tendencies in a manner remotely comprehensible to an amateur linguist like myself.



In this context, "Predicative Possession" is the perfect supplement to the WALS database. It looks in excellent detail at the question of predicative possession in a very good sample of languages (including quite a number not covered in Stassen's equally good Intransitive Predication) and aims to provide a clear understanding of why languages choose the particular type of possession encoding that they do. Before this, however, Stassen looks clearly at the various subtypes of each type of predicative possession so that one can understand how they evolve. "With" possessives especially are shown as likely to evolve into affixes that form adjectives, which creates two types depending on whether the language is tensed (and has nouny adjectives) or non-tensed (and has verb-like adjectives). in the most polysynthetic of languages such as Eskimoan or Wakashan, these tend to become quite like verbs, but Stassen shows that they cannot be considered verbs simply by virtue of the word structure of these languages.



In the next part of the book, Stassen looks for deciding characteristics to find out what determines the strategy used to determine what kind of possessive construction a language will have. He finds that two characteristics can determine what type of possessive construction a language will choose. These are whether its "when" clauses are "balanced" (ie. use the same verb forms as independent clauses) or "deranked" (use different verb forms from independent clauses) on one hand, and whether its nominal and locational predicates are "split" (different morphemes) or "shared" (as in English, with the same verb for both). He shows indeed that predicative possession structures in almost all languages are modelled on "when" clauses and that the underlying structure consists of the possessed item and an existential "be" verb. "Have" verbs form as a last resort when a language cannot distinguish a predicative possession sentence from a simple copular type ("is...").



All through "Predicative Possession" there is excellently done and very clear detail to illustrate the two key issues involved in determining how a language will encode predicative possession - so much so that none of the several hundred languages, even those which are extinct or not well-described, are omitted except in a few cases where there really does not exist adequate information. Consequently, "Predicative Possession" is a perfect complement to Stassen's other linguistics work "Intransitive Predication" and should be bought by all curious about linguistics.