Acquisition of Scalar Implicatures
|
Many studies have shown that young children do not derive scalar implicatures at an adult-like level, and children’s difficulties with scalar implicatures have been attributed to immature pragmatic capacities, the lack of lexical scales, or their limited processing capacities (e.g., Barner et al. 2011; Chierchia et al. 2001; Foppolo et al. 2020; Gotzner et al. 2020; Pouscoulous et al. 2007; Skordos & Papafargou 2016; Tieu et al. 2016). Yet, few studies have directly measured the role of processing limits, or comprehensively explored the predictions of processing-based accounts. This project aims to fill this gap by investigating processing accounts from three different perspectives. First, I test whether children’s working memory reliably predicts their performance when computing scalar implicatures. Second, I investigate whether a child performs better on scalar implicatures at the times when greater cognitive resources are available. Under a processing-based account, anything that facilitates the retrieval of alternatives should free up cognitive resources and thereby aid the child in computing scalar implicatures. Third, the project explores whether adults who are subjected to excess cognitive load will find it difficult to compute scalar implicatures. A processing account predicts that that they will. A number of previous studies have tested for an effect of memory load on adults’ computation of scalar implicatures (De Neys & Schaeken 2007; Marty & Chemla 2013; Pipijn 2014; Scrafton 2009; Van Tiel et al. 2019). Yet, the findings have been mixed. This project tries to clarify this picture by carefully considering the possibility that the effect of memory load may vary, depending on the scale.
|
Acquisition of Free Choice Inferences with dou-constructions
|
In Mandarin, a preverbal wh-phrase followed by dou ‘all’ is not a wh-question but a declarative statement with a free choice (FC) reading. Preverbal disjunction followed by dou also derives a FC reading. It has been suggested that the FC reading of these two constructions follows directly from the semantics of the particle dou (e.g., Xiang 2016, 2020). The FC reading of disjunction is prohibited in an episodic context or with a universal modal. Thus, Xiang (2020) argues for a covert exhaustifier in ‘disjunction + dou’ constructions. However, ‘wh-phrase + dou’ may sometimes (though not always) appear in an episodic context. Therefore, Xiang (2020) does not discuss whether there is any covert exhaustifier for ‘wh-phrase + dou’ constructions. In this case, evidence from child language can help the theoretical analysis.
On the view that the FC readings of both ‘wh-phrase + dou’ and ‘disjunction + dou’ follow directly from the semantics of dou, children should in principle acquire the two constructions around the same time. However, if there is a covert exhaustifier for ‘disjunction + dou’ but not ‘wh-phrase + dou’, it becomes possible for a child to acquire the former later than the latter, because the covert exhaustifier may complicate the learning process. My work aims to address this question by providing within-subject data from Mandarin-speaking children on the two dou-constructions. |
Acquisition of English Adjectival Resultatives
(together with Dr. Yasuhito Kido and Prof. William Snyder) |
Two distinctive types of complex predicates found in English are separable verb-particle combinations ('particles') and adjectival resultatives ('ARs'). Snyder (1995, et seq.) ties both to the positive setting of the Compounding Parameter ('TCP'). This predicts that during the acquisition of a [+TCP] language, any child who has acquired ARs or particles will also permit “creative” bare-stem, endocentric compounding. Existing support comes from children acquiring Japanese (Sugisaki & Isobe 2000) and English (e.g., Snyder 2007). Yet, the same evidence introduces two new puzzles: (i) why is compounding acquired roughly a year earlier in English than in Japanese? (ii) And in English, why is compounding always acquired at the same time as (and never substantially prior to) particles? Here we argue that both puzzles can be explained if we allow the trigger for a single parameter-setting (e.g., [+TCP]) to be completely different, for children acquiring different languages. Specifically, the trigger for [+TCP] (and hence, ARs) in English is proposed to be particles, which are unavailable in Japanese. Two novel predictions are tested and supported: (i) the frequency will be higher for particles than for any (other) potential trigger in child-directed English or Japanese; and (ii) children acquiring English (unlike Japanese) will have reliably adult-like comprehension of ARs by the age of three years.
|
Grammaticalization of yi 'one' in Mandarin
|
I have been investigating the grammaticalization of yi ‘one’ in Mandarin (overseer Prof. Zeljko Bošković), in light of cross-linguistic evidence (e.g., Cantonese, Turkish, and Slovenian). I propose that due to an ongoing grammaticalization process, Mandarin yi is ambiguous between a numeral and an indefinite article. The grammaticalization of yi is an instance of a well-known phenomenon: specifier-to-head reanalysis (e.g., van Geldern 2001, 2004; Roberts 1993; Roberts & Roussou 2003). Yet, on the basis of cross-linguistic evidence from Cantonese and Turkish, I argue for the existence of an intermediate stage in specifier-to-head reanalysis: a functional category starts out in the specifier of a phrase; it then adjoins to another head; and finally it is reanalyzed as an independent head projecting a separate phrase.
|
Adjectives in American Sign Language (ASL)
|
In the area of sign linguistics, I am involved in research addressing theoretical questions about American Sign Language (ASL). In particular, I have been examining the properties of ASL adjectives (under the guidance of Prof. Diane Lillo-Martin). My work aims to comprehensively investigate ASL adjectives and provide a unified account for a number of their characteristics. I argue that: 1) prenominal adjectives are adjuncts to the noun, and follow the ordering constraints proposed by Cinque (1994); 2) post-nominal adjectives are predicates of relative clauses; and 3) ‘adjectives’ that can be used with aspectual inflections are in fact verbs, as proposed by Bernath (2009).
|