Katerina Stathi: Granularity in the verbalization of events and objects

This study is concerned with differences in granularity, i.e. the choice between semantically general (coarse-grained) and specific (fine-grained) words, with data from 13 conceptual domains.The research question is formulated in Chapter 1 'Introduction' in a very general way: "Do languages verbalize states of affairs at a similar level of detail consistently?Is there a preferred level of detail in expressing meaning?" (p.13).Then, Chapter 2 introduces the theoretical background, Chapter 3 describes the methodology, and each of Chapters 4-15 is devoted to one of the conceptual fields that are studied.These chapters are structured in a similar way, which makes the reading of this comprehensive work easier.The results are then summarized and theoretically discussed in Chapter 16, followed by a conclusion in Chapter 17.The book is based on data elicited with visual stimuli.In lexical analysis, elicitation with video clips or pictures supports an analysis made from a world-toword (Malt et al. 2010) or onomasiological (Geeraerts 2010) perspective, starting with a conceptual structure and looking at its various lexico-syntactic realizations.As I have studied several of the domains (referred to as 'semantic fields') from a semasiological perspective based on corpus data, in what follows I will compare both perspectives, which will make the review more subjective but hopefully more interesting.Corpus-linguistic methods start with examples of wordsoften in a sentence contextand support the study of words from a word-to-world (or semasiological) perspective.This approach makes it possible to study how the use of a certain word in a specific context is related to the complete set of uses of the word, i.e. its meaning potential.After a review of the theoretical background (domains and fields, theories of lexicalization), a very precise version is presented of the 'granularity hypothesis' in Chapter 2 (pp.66-67): "Satellite-framed languages (S-languages) are semantically more specific than verb-framed languages (V-languages)".The hypothesis was tested on very extensive data from four languages: English and German (S-languages), and Greek and Turkish (V-languages).The study covers 10 conceptual domains of verbs and three domains of nouns.