Mirror Lake Papers
, Nilufer Sener, Carlos Buesa García and Tsuyoshi Sawada 2011
Ana Bastos-Gee, University of Connecticut
On Exclamatives with a Bothering Inference
In this paper, I study four types of exclamative constructions in Brazilian Portuguese (BP), characterized by a negative-bias inference and propose a unified analysis for them by adapting the alternative semantics for focus, proposed by Rooth (1992), in which given a question-answer pair (?,a), focus is interpreted at the level of a, using the Focus Interpretation Principle. Constructions with bothering inferences are not used to ìanswerî questions, i.e. the set of propositions that provides an appropriate antecedent for the focus variable does not come from a question. I propose that the set of propositions is pragmatically established on the basis of speakersí expectancies and attitude, and that these constructions involve a specialized two-place exclamative operator, the BOTHER operator. The operator carries a presupposition, which requires the proposition of this type of exclamative utterances to be less expected and less approved in the set of propositions pragmatically established by the speaker.
Harry van der Hulst, University of Connecticut
In this article, I will discuss some aspects of so-called constraint-based phonological theories. Of central concern is the notion phonological level and I propose that at least two levels must be adopted, the I-level (idealized, lexical representations) and the U-level (utterances). The latter level forms the input to the phonetic implementations module. Then, I turn to a discussion of the roles of constraints, parameters and repair rules, arguing against the notion of constraint ranking. I will synthesize my conclusions in describing the approach called Head-driven Phonology (HDP) and apply this model to a set of well-known Yawelmani data which involve vowel shortening, vowel epenthesis and vowel harmony.
Serkan Sener, University of Connecticut
The Acquisition of Causatives in Turkish
This study investigates the acquisitional predictions of a cross-linguistic generalization due to Baker (1988), which states that if the transitive causative construction involves a special case marker, rather than Accusative, on one of the internal arguments, the simple/underived ditransitives will also require a special Case on the Goal argument. Taking adult Turkish as a case in point, which shows compatibility with Bakerís generalization, the present study inquires into the question of whether this generalization is supported by evidence from language acquisition. To this end, an idea that is strongly argued for in Snyder and Stromswold (1997), Snyder and Sugisaki (2002), and Snyder (2007) is made use of, which takes ëtime course of language acquisitioní as an important factor in understanding child language acquisition. The way the generalization was stated in Baker (1988), a one-way implication, predicts an ordered acquisition of the two grammatical constructions under investigation. This prediction is borne out as the results of the investigation show that the knowledge required for underived ditransitives forms a subset of the knowledge required for the acquisition of transitive causatives, at least in Turkish.
Takuro Tanaka, University of Connecticut
This paper develops an analysis to investigate a condition for so-called floating numeral quantifiers (FNQs) in Japanese. Cross-linguistic differences for FNQs have been discussed in previous literature. FNQs in Japanese, however, show idiosyncratic acceptability which previous studies cannot explain. Miyagawa (1989) suggests mutual c-command requirement to explain such controversial data of Japanese FNQs. This paper provides Japanese data which Miyagawaís analysis cannot cover, and suggest that a crucial condition for Japanese FNQs is aspectual delimitedness. Data with individual-level and state-level distinction, unergative and transitive verb, progressive, and psych verb provide unpredictable contrast for mutual c-command analysis. Based on such data, I argue that FNQs in Japanese are an aspectual-sensitive phenomenon, and they are licensed by the delimited property assuming a particular temporal endpoint or limited number of individuals.
Jon Gajewski, University of Connecticut
Implicature Projections: Comments on Chierchia
This paper, originally written in 2001, addresses the issue of how to calculate the scalar implicatures of a sentence that contains scalar items embedded under other operators, scalar and not. The main proposal is a reaction to the proposal made in Chierchia 2001. Chierchia introduces a mechanism that generates a set of strengthened meanings alongside the plain truth-conditional meanings of a sentence. The current proposal arose out of a criticism of Chierchiaís original formulation, which was later amended in Chierchia 2004. I propose to assign to each a scalar item a pair of a plain meaning and a strengthened meaning. A compositional algorithm then assigns to each constituent a plain meaning and a unique strengthened. The predictions of the theory match Chierchiaís in most cases, but differs with it in its predictions under DE quantifiers. The current proposal predicts strengthened meanings that are likely too strong.