Whamit!

The Weekly Newsletter of MIT Linguistics

Issue of Monday, May 17th, 2021

Phonology Circle 5/17 - Dóra Kata Takács (MIT)

Speaker: Dóra Kata Takács (MIT)
Title: Similarity-based syncretism in a Hungarian dialect
Time: Monday, May 17th, 5pm - 6:30pm

Abstract: In this talk I take a closer look at the indicative definite and the imperative forms of t-final verbs in Hungarian. In standard Hungarian t-final verbs undergo a special phonological process in the imperative when they are followed by a j-initial suffix (cf. Siptár and Törkenczy 2000). Interestingly, there is a substandard very stigmatized version of Hungarian where this phonological process extends to the entire verbal domain, but crucially not beyond that. I present an analysis for the verb forms in standard Hungarian that not only provides an insight into why t-final verbs are special, but can also successfully capture the changes other verbs undergo. Furthermore, I argue that the dialectal variation is best explained as similarity-based syncretism following Steriade (2016).

Syntax Square 5/18 - Shigeru Miyagawa (MIT), Despina Oikonomou (HU-Berlin)

Speaker: Shigeru Miyagawa (MIT), Despina Oikonomou (HU-Berlin)
Title: Case and A/A’-movements
Time: Tuesday, May 18th, 1pm - 2pm

Abstract: We argue for a new way to look at the constellation of data having to do with Condition C reconstruction effects. We propose that Case, which is a key component of Wholesale Late Merger (Takahashi 2006, Takahashi & Hulsey 2009 (TTH)), is an issue not about the head of the chain, as TTH argue, but rather it should be viewed as an issue associated with the lower copy:

Case Interpretation Requirement (CIR)
Case-marked copy of movement must be interpreted in full.

This accounts for the familiar cases of Condition C reconstruction effects reported in the literature. It also accounts for A-movement scrambling that unexpectedly triggers a Condition C reconstruction effect (Hindi, Bhatt and Keine 2019; Japanese, Saito 1992; Mongolian, Gong 2021). Here A-movement behaves like A’-movement because the lower copy is case-marked. Similar facts hold for the CLLD construction in some Romance languages (e.g., Cecchetto 2000).

We also argue that the CIR accounts for a decay in Condition C effect in long-distance wh-movement of DPs (Adger, et al. 2017; Bruening and Al-Khalaf 2017).

(1) ?*[Which criticism of Johni] did hei deny___?
(2) (?)[Which criticism of Johni] does hei think [CP that Mary believes___]?

Some linguists have argued that this decay indicates that there is no Condition C reconstruction effect in wh-DP-movement. We will argue instead that, due to the PIC, the LD-moved wh-phrase only has access to the intermediate copy in the lower Spec,CP (below he in (2)), which is a non-Case position. By the CIR, this intermediate copy need not be (fully) interpreted, thus leading to a decay in Condition C effect. In this way, the intermediate copy of wh-movement behaves similarly to the copy of A-movement. We will demonstrate that in languages in which the intermediate copy is case-marked (e.g., Greek, Japanese), there is no decay in Condition C effect whatsoever regardless of the distance between the LD-moved element and the pronoun.

LF Reading Group 5/19 - Nina Haslinger (University of Göttingen)

Speaker: Nina Haslinger (University of Göttingen)
Title: Cumulativity as global non-maximality: A non-lexical underspecification approach
Time: Wednesday, May 19th, 1pm - 2pm

Abstract: In this talk, I argue that cumulative construals of sentences with multiple plurals (e.g. a construal of (1) that only requires the existence of a plurality of two students such that each of them read books and each book was read by at least one of them) are subject to contextual constraints that parallel those for non-maximal construals of plural definites (e.g. a construal of (2) on which not all of the windows have to be open). In particular, the QUD-dependence observed for non-maximal construals (Malamud 2012, Križ 2015 a.o.) and their sensitivity to contextual alternatives of the predicate (Križ 2015) carry over to cumulativity.

(1) Two of my students have read all the books. (2) The windows are open.

Since cumulativity does not literally reduce to non-maximality, a unified approach to both phenomena is needed. Existing arguments against an ambiguity approach to non-maximality can be extended to argue that (1) should not have separate LFs for the distributive and the cumulative reading (cf. Schwarzschild 1996, Kratzer 2008). In simple cases like (1), the context-dependency behavior of cumulativity can be accounted for by a straightforward extension of the three-valued approach to non-maximality, so that (1) expresses a three-valued proposition that is neither true nor false in a cumulative scenario.

I then show that this approach will not extend to all cases of cumulativity, based on German data involving plurals within an embedded clausal conjunction (cf. Schmitt 2019). The problematic cases are analyzed within a variant of Schmitt’s (2019) system in which any expression containing a plural denotes a predicate of pluralities (possibly of higher type). To account for the context-dependence of cumulative and non-maximal construals, I propose that these predicates of pluralities should be three-valued, with ‘classical’ and ‘tolerant’ extensions that are computed in parallel. Plural sentences end up denoting a three-valued predicate of propositions, which leads to a gap between the truth conditions and the falsity conditions that is resolved in a QUD-dependent manner. I conclude by outlining some potential advantages of this system over the standard ambiguity approach to cumulative and distributive construals.

MorPhun 5/19 - Boer Fu (MIT) and Danfeng Wu (MIT)

Speaker: Boer Fu (MIT) and Danfeng Wu (MIT)
Title: Numeral Allomorphy in Mandarin Chinese and A Synthetic Side of the Language
Time: Wednesday, May 19th, 5pm - 6:30pm

Abstract: In Mandarin Chinese, the numerals 1 and 2 display interesting allomorphy patterns between two distinct forms. We discuss two puzzles involving single-digit and multi-digit cardinals, their forms when immediately followed by a classifier, as well as their ordinal forms. We argue that we need a morphological rule that refers to both linear order and structural / prosodic relations. Furthermore, the different morphological patterns of numbers below 100 vs. numbers above 100 suggest that Mandarin, despite lacking a synthetic vs. analytic distinction on the surface, in fact makes such a distinction in ways parallel to English. Like English, Mandarin numbers under 100 are synthetic (parallel to English fif+teen and fif+ty), while bases of hundreds and above are analytic (parallel to English five#hundred). We further observe a monotonic trend across languages, where lower number bases tend to be irregular (e.g. suppletive and readjusted), and higher number bases tend to be regular (e.g. analytic). This observation, if real, is interesting because it can be described as *ABA, but cannot be straightforwardly explained by the containment analysis, the traditional analysis of *ABA in the literature.

LingLunch 5/20 - Norvin Richards (MIT)

Speaker: Norvin Richards (MIT)
Title: Finding something to lean on: some extensions of Affix Support
Time: Thursday, May 20th, 12:30pm - 1:50pm

Abstract: I’ll present some ongoing research that attempts to account in a unified way for the FOFC (and some of its exceptions), the Head-final filter, AP-adjacency (Giurgea 2009, Adger 2012, Belk and Neeleman 2014), and some conditions on extraposition in Hebrew discussed by Sichel (2002). The account is based on Contiguity Theory, and will involve some tweaks to the existing idea of Affix Support; the offending configurations will all involve a head in need of Support which fails to find it, once the condition on Affix Support is properly understood.