Scalar Meanings of the Concessive (-to), the Contrastive Topic Marker (-nun) and -man ‘only’ in Korean (and Japanese)

Scalar Meanings of the Concessive (–to), the Contrastive Topic Marker (-nun) and –man ‘only’ in Korean (and Japanese) Chungmin Lee Seoul Nat’l University Dept of :Linguistics and Cognitive Science Program, Seoul, Korea clee@snu.ac.kr Abstract This paper investigates the systematic relatedness between the concessive –to (or mo)-marked polarity phenomenon and the high tone –nun (or wa)-marked Contrastive Topic phenomenon with respect to underlying concessivity and thereby derived scalarity. I also investigate the scalar features of the exhaustivity focus marker –man (or dake) in inherently scalar numeral, quantifier and predicate contexts. I further try to show a relation between polarity/negativity and implicature (suspension). 1 Introduction The concessive –to (or mo)-marked (attached to an Indefinite) polarity phenomenon and the high tone –nun (or wa)-marked Contrastive Topic (CT) phenomenon show a close relatedness, with respect to underlying concessivity and thereby derived scalarity. The scalar but not concessive features of the exhaustivity focus marker –man (or dake) in inherently scalar numeral, quantifier and predicate contexts will also be described. Some relatedness between polarity/negativity and implicature (suspension) will be investigated. Negative CT Ss invoke affirmative weaker implicatures connected by a PA ‘but’ and metalinguistic negation constitutes a Contrastive Focus construction. I argue that both polarity, based on –to ‘even,’ and CT, -nun with a high tone, involve conventional scalar implicatures motivated by concessivity. 2 Polarity: Concessivity and Scalarity 2.1 Where Does Scalarity Come from? When the concessive marker –to (in K)/–mo (in J) is attached to an indefinite such as a numeral/minimizer and INDefinite-wh-/ amu- (and any with implicit even (Y. Lee and Horn 1994)) (C. Lee 1993, 1996), it generates an NPI, as in (1) and (3): (1) a. han saram –to an w-ass-ta one person-CNC not come-PAST-DEC b. hitori -mo ko-nakatta (a, b) ‘One person-CNC didn’t come.’ (2) a. han saram –i[NOM] an w-ass-ta (NOM=Nominative) b. hitori (-ga) ko-nakatta (a, b) ‘One person didn’t come.’ (3) a. *han saram –to w-ass-ta b. *hitori -mo kita (a, b) ‘*Even one person came.’ (cf. Not even one person came.) In (1), han saram ‘one person’ is indefinite, nonspecific and not a partition requires a negation as an NPI, if it is followed by the Concessive (CNC) marker, reaching universal negation by reversed entailment. If PACLIC 18, December 8th-10th, 2004, Waseda University, Tokyo - 217 - it is followed by a NOM, as in (2), however, it must be either specific or a partition, though not definite, taking wide scope over the negation as an existential, in the S. So, 'more didn’t come' is not scalarly entailed by (2). Attributing polarity simply to focus, as done by Rooth (1985), Krifka (1994) and partly Lahiri (1998) or to simple scalar implicature suspension, as proposed by Chierchia (2002) is not adequate enough. Chierchia's approach simply lacks an ultimate motivation behind polarity. Concession triggers an adversity scale and suspends scalar implicatures. So, contrary to his expectation, an implicature of 'not both' is suspended in the following S because of the concessive 'even' without any negative predicate such as 'doubt': (4) …

Content truncated to 500 words