Zheng Shen 申正

D-linking effects in wh-fronting and wh-in situ

Villata and Sprouse 2023 show a partial amelioration effect of d-linked wh-elements across certain islands (e.g. whether island). Our experiments from Colloquial Singapore English reveals two types of cross-linguistics variation. D-linked wh-elements in Singlish wh-fronting questions improve sentences with long distance movement regardless of island, while d-linked wh-elments in Singlish wh-in situ questions do not have any effects. We argue that the d-linking effects can have two sources. The partial feature overlap analysis in the tradition of fRM can be observed in English but the C in Singlish is only specified with [+Q] and not [+Q, +N]. Thus partial amelioration is only seen in English. On the other hand, d-linked wh-elements make processing long distance dependency easier (Hofmeister and Sag 2010) which is observed in Singlish (and probably also present in English). Wh-in situ questions do not involve long distance dependency, thus no d-linking effect is observed.

Shen, Zheng. and Beth Chan. Wh-island effects and d-linking effects in wh-in situ questions. HSP 37 | WCCFL 42


Extraction out of NPs: Definitness island effect

Extraction out of the nominal domains (NP/DP) is sensitive to multiple factors, many of which have a hard time finding their places in syntax. This project uses experimental methods to 1. build a solid empirical foundation and 2. tease the existing locality theories apart.

Using factorial design comparing extraction out of DPs headed by indefinite and definite articles (a vs. the), we found significant island effect in wh-questions with bare wh-words (Exp Set 1). The effects become marginal when using a D-linked wh element (which) in wh-question and in relative clauses (Exp Set 2).

Shen, Zheng and Meghan Lim. 2021. Extraction from definite, indefinite, and superlative NPs: An experimental approach. NELS 52, Rutgers University (online).

However, when comparing demonstratives (that) with indefinite (a/an), both D-link wh-questions and relative clauses show significant island effect again (Exp Set 3). In two follow up experiments comparing that and the indefinite article in D-linked wh-questions and relative clauses (Exp Set 4), we used sentences with piped-piping (instead of P-stranding in Set 1-3) since previous studies observe difference in islandhood between sentences with piped-piping and P-stranding. We found the same island effects as Exp Set 3.

Shen, Zheng and Meghan Lim. accepted. The definite DP island in wh-questions and relative clauses. Syntax. (presented at HSP 2022 | WCCFL 40)

Nick Huang and I found definite island effects in both wh-fronting in English and wh-in situ in Chinese, however, the island effects are sensitive to verb of creation in English and not in Chinese. We argue that this pattern supports a combination of the PIC and the Specificity Condition to account for the definite island. I tested indefinite NPs with possessors in Colloquial Singapore English and the results from wh-fronting and wh-in situ in CSE support the PIC+Specificity approach.

Shen, Zheng and Nick Huang. in revision. The role of phases and specificity in definite islands.

Shen, Zheng. in prep. Both PIC and Specificity block wh-dependency out of NPs: Evidence from Colloquial Singapore English.


Coordinate Structure Constraint violating movement and conjunction agreement

This projects investigate movement that violates Coordinate Structure Constraint and its effect on conjunction agreement in Bosnian-Croatian-Serbian. Agreement with conjunction can probe the highest conjunct, the linearly closest conjunct, or default masculine agreement. I show that once the first conjunct moves across the agreement target, it becomes the only possible agreement controller, other agreement options become unavailable. I argue that this can be accounted for in the non-linear approach to conjunction agreement but poses difficulties for the linear appraoch.

Shen, Zheng. 2023. Conjunction Agreement and the Coordinate Structure Constraint, Glossa: a journal of general linguistics 46(1). DOI: https://doi.org/10.16995/glossa.6382

On the backburner

Multi-valuation across languages and domains

Multi-valuation involves one probe being valued by multiple goals, thus receiving multiple values. I look at when one probe receivs two singular values. In the nominal domain, this is illustrated in Nominal Right Node Raising constructions. Two agreement patterns are observed: the Distributive agreement where the multi-valued probe (noun) must be singular (e.g. English Slovenian), the Summative agreement where the multi-valued probe must be plural (Bulgarian, Russian). IN the TP domain, multi-valuation is showcased in T' RNR construction reported in Groß 2015. The Distribution agreement is observed in Slovenian while the Summative agreement is observed in English. I survey around 20 languages and observe a 3/4 pattern regarding the two agreement patterns in two different domains. I propose that multi-valuation shows the Agreement Hierarchy effect proposed for hybrid noun agreement (Corbett 1979 a.o.).

Shen, Zheng. 2019. The multi-valuation Agreement Hierarchy. Glossa: A Journal of General Linguistics, 4(1), 46. DOI: 10.5334/gjgl.585

The N and T in NP RNR and TP RNR are multi-dominated, thus multi-valued.

The shared N in NP RNR and share T in TP RNR can show distributive or summative agreement.

Based on cross-linguistic survey, no language show summative agreement on N but distributive agreement on T. This pattern is parallel with the Agreement Hierarchy observed for morphological and semantic agreement in hybrid noun agreement.

Summative agreement in multi-valuation is semantic agreement and distributive agreement in multi-valuation is morphological agreement.

Shen, Zheng. 2018. Feature Arithmetic in the Nominal Domain. Doctoral dissertations, University of Connecticut.



Morphological and Semantics Agreement

Cross-linguistically, hybrid nouns like committee can control either morphological agreement and semantic agreement: The committee has/have met. This optionality has motivated a feature theory with both semantic features and former/morphological features. Although formal features have been investigated extensively in both syntax and morphology, the nature of semantic features remains unclear. Multi-valuation provides a useful probe into semantic features. Grosz 2015 argues that semantic number features are referential indexes. However Belk and Neeleman 2018 proposes that semantic agreement results from a repair rule when syntactic agreement is blocked. We show that both a repair approach and a index approach are problematic but a particular combination of the two is descriptively adequate. Currently we are working on extending the empirical scope of semantic agreement to switch reference observed in indigenous languages in North and South America.

Shen, Zheng and Peter W. Smith. 2019. Morphological and semantic agreement beyond hybrid nouns. Proceedings of NELS 49.

Shen, Zheng and Peter W. Smith. 2018. Morphological and Semantic Agreement with Multiple Values. Multiple Agreement across Domains 2018, Berlin, Nov. 8-9, 2018.



Acquisition of Superlatives

Previous acquisition literature on superlatives focuses on the emergence of superlative morphology in children's production. Lyn Tieu and I have conducted corpus and experimental studies to probe children's knowledge of the different interpretations of superlative expressions. Previous studies report children's difficulties in interpreting superlative expressions as adults do. We found that Question under Discussion (QuD) to be crucial in probing children's ability to interpret ambiguous expressions. Once the QuD is made salient, children largely behave like adults in interpreting "the largest cake by Monkey". However, there is some difference between interpreting fragment answers and full answers.

Tieu, Lyn and Zheng Shen. 2022. Children’s Interpretation of Superlatives in Full and Fragment Answers. Proceedings of BUCLD 46. (lingbuzz link)

Tieu, Lyn and Zheng Shen. 2018. Interpretive restrictions on superlatives in full vs. fragment answers. GLOW 41, Budapest, April 2018.

Tieu, Lyn and Zheng Shen. 2017. Aiming high: children’s knowledge of absolute and relative readings of superlatives. GALA 13, Sept. 2017.

Tieu, Lyn and Zheng Shen. 2016. “The littlest linguists and their superlatives: a first experiment.” Proceedings of CLS 50.

Tieu, Lyn and Zheng Shen. 2015. “Searching for absolute and relative readings of superlatives: a second experiment.” Proceedings of BUCLD 39.


Fragment Answer and Superlatives

A number of approaches have been proposed to derive fragment answers. Most of the arguments center on the mobility and the island sensitivity of the fragment. This project takes a different route and teases apart the various approaches by examining the interpretative differences between the fragment answer and the full answer. In superlatives in English, the availability of the relative reading with NP internal focus (RIN) is facilitated by overt movement and not covert movement, which provides a test case regarding whether fragment answers involve overt movement. The fact that fragment answers, but not their full-sentence counterparts, allow the RIN shows that at least some fragment answers involve overt movement in narrow syntax.

Shen, Zheng. 2018. "Fragment answers and movement - a superlative argument, Natural Language and Linguistic Theory, DOI: 10.1007/s11049-017-9369-9.



Agreement in Nominal Right Node Raising

Nominal right node raising constructions involve two DPs sharing an constituent, e.g. "this tall and that short student". The number marking on the shared element ("student") depends on the type of the sharing elements: "this tall and that short student(*s) are a couple", "John's and Mary's student*(s) are a couple". I conduct a cross-linguistic survey and show that the key factor lies in whether the sharing elements agree with the shared element. I provide a multi-dominance analysis for the construction and argue that when a noun is valued with two instances of singular features, it is spelled out as singular rather than plural or dual, unlike verbal elements in TP RNR (Grosz 2015) and local effects (Gluckman 2015).

Shen, Zheng. 2015. No Clash Constraint in Nominal RNR Number Agreement in Proceedings of PLC 39.



Relative Reading with NP Internal Focus of Superlatives (RIN)

In addition to the absolute reading and the relative reading of superlative expressions like "the best album by U2", Pancheva and Tomaszewicz (2013) observe a third reading that involves an NP internal focus (RIN) in Polish and Bulgarian. I extend the empirical scope of the inquiry in both Polish-type languages and English-type languages, showing that overt, but not covert movement, facilitates the availability of the third reading. I argue that the generalization is a conspiracy of semantic requirements and syntactic locality constraints.

Shen, Zheng. 2014. The Third Reading of the most expensive photo of Abby in Proceedings of PLC 38.

Shen, Zheng. 2013. On the Relative Reading with NP Internal Focus in Superlatives in Proceedings of WCCFL 31.