Two Types of Nominalization in Japanese as an Outcome of Semantic Tree Growth

Copyright 2012 by Tohru Seraku 26th Pacific Asia Conference on Language,Information and Computation pages 153–162 Two Types of Nominalization in Japanese as an Outcome of Semantic Tree Growth Tohru Seraku St. Catherine’s College, University of Oxford Manor Road, Oxford, UK. OX1 3UJ tohru.seraku@stcatz.ox.ac.uk Abstract The particle no in Japanese exhibits two types of nominalization: “participant” and “situation” nominalization. Despite several motivations for a uniform account, only a few attempts have been made to address no-nominalization uniformly. In this paper, I shall develop a unified account within the formalism Dynamic Syntax, and show that a number of properties of the phenomenon follow from the analysis. 1 Introduction The particle no in Japanese displays two types of nominalization: “participant” nominalization (1) and “situation” nominalization (2). (1) [Akai no]-o Tom-ga nagu-tta. [red NO]-ACC Tom-NOM hit-PAST ‘Tom hit a/the red one.’ (2) [Mary-ga kireina no]-o [Mary-NOM beautiful NO]-ACC Tom-ga shi-tteiru. Tom-NOM know-PRES ‘Tom knows that Mary is beautiful.’ In participant nominalization, the particle no turns a preceding clause into a nominal that denotes an object or a person. In situation nominalization, the particle no turns a preceding clause into a nominal that denotes an event or a proposition. A case of ambiguity is presented in (3). (3) [Nai-ta no]-o Tom-ga mi-ta. [cry-PAST NO]-ACC Tom-NOM see-PAST a. ‘Tom saw someone who cried.’ b. ‘Tom saw the event of someone’s having cried.’ Participant nominalization is exemplified by (3a), and situation nominalization by (3b).1 One issue that immediately arises is whether no in (1, 2, 3) should be treated uniformly. In other words, does no in (1, 2, 3) form a single item or are there two nos one of which appears in (1, 3a) and the other of which appears in (2, 3b)? Seraku (in press) defends a uniform analysis based on several motivations (e.g. methodological, cross-linguistic, functional, diachronic). Despite these motivations, a unified analysis of no has been largely untouched (e.g. Kitagawa, 2005; Kitagawa and Ross, 1982; Murasugi, 1991; Shibatani, 2009; Tonoike, 1990). Against this background, the aim of the present paper is twofold as follows. First, I shall articulate a unified analysis of no-nominalization within the grammar formalism Dynamic Syntax (Cann et al., 2005; Kempson et al., 2001). Second, I shall show 1 Seraku (in press) summarizes diachronic data that give credence to the exclusion of such data as (i) from the analysis to be developed in this paper. (i) Tom-no Tom-NO ‘Tom’s’ 153 that the analysis captures a range of characteristics of the phenomenon. 2 Dynamic Syntax Dynamic Syntax (DS) is a formalism that models “knowledge of language”, which is conceived as a set of constraints on language use (Cann et al., 2005; Kempson et al, 2001). Language use consists of production and comprehension. DS is shown to model production (Cann et al., 2007; Purver et al., 2006), but this paper focuses on comprehension. DS is then said to provide a set of constraints on how a parser builds up an interpretation gradually as it processes a string word-by-word online. DS models gradual …

Content truncated to 500 words