Twenty years ago now, my book Parts and Wholes in Semantics was published. It is a systematic application of gestalt principles (conditions of integrity) to a core area of semantics, the mass-count distinctions and a range of part-whole-related expressions and constructions in natural language. As such, it develops an alternative to the still dominant extensional mereological approach (of which it gives conceptual and empirical critique). Today, I think gestalt principles apply in a more construction-oriented way rather than the information-given way pursued in the book, and I now believe in plural reference (rather than referenve in a plurality), see my paper Plural Reference and Reference to a Plurality. But the book is at least as relevant now as then, both regarding its conceptual content and its empirical generalization, especially within the perspective  of natural language ontology that I am developing. Here my OUP blog entry of this year entry for the book.