university of california, berkeley


The Circle is a weekly forum dedicated to discussion of the descriptive, experimental, and theoretical study of syntax and semantics, featuring presentations of ongoing research by members of the Berkeley Linguistics Department and other departments, as well as discussion of previously published works.


Fridays, 3:00-4:30pm


Room 1303 (Dwinelle)

Recurring link.


Shweta Akolkar

Kang 'Franco' Liu


University of California, Berkeley
Department of Linguistics

Fall 2023


3 may
Ekaterina Georgieva (Hungarian Research Centre for Linguistics/Stanford)
On sluicing in Meadow Mari and Udmurt

This talk zooms in on sluicing in Meadow Mari and Udmurt (Uralic). The aim of the talk is to first examine whether the elliptical sentences employed in these two languages show the footprint of isomorphic sluicing, where the elided material is structurally identical to the antecedent clause, or are rather instances of non-isomorphic sluicing derived from e.g. copular clauses or clefts. It will be shown that sluicing in Meadow Mari and Udmurt does show properties associated with isomorphic sluicing. Additionally, what makes the Meadow Mari and Udmurt data interesting is that wh-remnants can optionally bear possessive morphology; it will be examined what its role is. The talk will then explore possible accounts that capture the properties of sluicing in the two languages.

26 april
Round robin!

19 april
Justin Royer & Tessa Scott (UC Berkeley)
Directionals and argument structure in Mam

In this paper, we propose that most verbs in Mam (Mayan) are semantically underspecified: they lack argument structure beyond taking an event argument. That is, not only are external arguments (agents) severed from the semantic content of Mam verbs (à la Kratzer 1996), but so are themes. Instead, we propose that the dedicated function of introducing themes falls not to the verb itself, but directional auxiliary verbs (“directionals”), which must accompany the majority of verbs that require themes (i.e., unaccusatives and transitives). The analysis builds on previous work on Mayan directionals, which has also connected their function to argument structure (Mateo-Toledo 2008, 2023). (This is joint work with Noah Elkins)

12 april
No meeting. We are hosting WCCFL 42!

5 april
No meeting.

29 march
No meeting. Happy spring break & Happy Cesar Chavez Day!

22 march
Josh Phillips (Stanford)
Djambarrpuyŋu morphosemantics: tense, reality status & the verbal paradigm

In Djambarrpuyŋu—a Yolŋu language spoken in northern Australia—verbs stems are inflected for one of four different inflectional categories. Previous descriptions of these categories’ functional domains have eschewed a unified analysis of their semantics. The reason for this is the apparent polyfunctionality of each inflection, in large part driven by a number of striking phenomena, notably: (1) cyclic tense (the temporal interval/s that are compatible with each inflection are ostensibly discontinuous) and (2) asymmetric negation (where a mood/reality status distinction available in positive clauses is unavailable in negative ones.) In this talk, I provide a branching-times-theoretic analysis of of Djambarrpuyŋu’s inflectional categories as tense/mood markers by appealing to two privative semantic features: a temporal one (precontemporaneity) and a modal one (nonveridicality).

21 march
András Bárány (Bielefeld University)
Agreement and the syntax–information structure interface

There are different analyses of how syntax and information structure (IS) interact. Some approaches assume that IS features are present in narrow syntax, while others ban IS features from syntax altogether. Some of these approaches (e.g. Reinhart 2006, Neeleman & van de Koot 2008, Titov 2020) assume that syntax freely generates structures which are then matched to discourse templates at a post-syntactic interface to information structure.

In this talk, I discuss agreement phenomena that appear to be conditioned by information structure. These agreement patterns are more difficult to analyse using interface filters as agreement is not generally assumed to apply as freely as movement. Instead, I propose that agreement sensitive to information structure can be modelled using narrow syntactic information structural features in a way parallel to the analysis of agreement patterns sensitive to person features of multiple arguments.

15 march
Amy Rose Deal (UC Berkeley)
Probe specific locality

Syntactic dependencies consistently show locality constraints, but these constraints have traditionally been taken to vary from one type of dependency to another (e.g. A' but not A movement can leave a finite clause) as well as from one language to another (e.g. out of finite clauses, some languages allow A movement, and some languages disallow A' movement). In this talk I will sketch a perspective on locality effects in syntax that attempts to find underlying unity in this domain while accounting for this diversity of behaviors. This is part of a larger project that seeks to make explicit a single operation of Agree that can account for all long-distance dependencies in syntax—the interaction/satisfaction theory. Special attention will be paid to the difference between 'single' and 'multiple' Agree, movement with mixed A/A' characteristics, and patterns of selective opacity beyond the A/A' distinction.

8 march
Rebecca Jarvis (UC Berkeley)
Long-distance obligatory binding: Distinguished variables and disjointness requirements

The notion of obligatory binding is well-known from the analysis of elements that are subject to Condition A. However, not all obligatorily-bound items obey Condition A: some items, like logophoric pronouns, are subject to obligatory-binding constraints that are more long-distance in nature. In this talk, I focus on a case study of pronouns in Atchan (Kwa, Côte d'Ivoire) to argue that such long-distance obligatory-binding requirements are fundamentally semantic in nature. In particular, I propose that certain Atchan pronouns bear presuppositional disjointness conditions that restrict their (co)referential and binding possibilities. On this view, obligatory-binding phenomena involve a pairing of (a) a presuppositional identity feature and (b) a specific binding operator that binds a variable in that identity feature. I argue that the analysis I propose in Atchan can be extended in a meaningful way to other obligatory-binding phenomena, including Algonquian-style obviation systems (Branigan & Mackenzie 1999, Oxford 2017), logophoric pronouns (von Stechow 2003, Pearson 2015, Bassi et al. 2023), and so-called anti-logophoric pronouns (Hill 1995, Andersen 1999, a.o.).

1 march
Wendy Liz Arbey López Márquez (UC Berkeley)
The structure of Nuntajɨɨyi DP and non-DP relative clauses

In this talk I will discuss the structure of relative clauses in Nuntajɨɨyi (also known as Highland Popoluca or Sierra Popoluca), an understudied language from the Zoquean branch of the Mixe-Zoque family, spoken in southern Veracruz, Mexico. I show that the language exhibits surface-level differences among two types of relative clauses: “DP relative clauses”, whose heads are DPs, and “non-DP relative clauses”, whose heads are non-DPs (e.g., oblique heads). These two types of relative clauses use different relativizing morphemes, as is the case in other languages (see e.g., Hsieh 2020 on Tagalog), and are not always compatible with the same complementizers. Despite their surface structural differences, I argue that both DP and non-DP relative clauses involve Ā-movement, as they form long distance relative dependencies and are sensitive to island constraints. Relying on reconstruction and extraposition diagnostics, I propose that DP relative clauses not only have properties associated with a true head-external structure (Partee 1975, Chomsky 1977), but also have properties associated with a head-raising structure (Schachter 1973, Vergnaud 1974, Kayne 1994, Sauerland 1998, Bianchi 2000, Bhatt 2002, Bodomo & Hiraiwa 2010, Hiraiwa 2017, whereas non-DP relative clauses only have properties associated with a true head-external structure. To account for the different structures, I adopt Cable’s (2007, 2010) Q-based theory and assume that Nuntajɨɨyi DP and non-DP relative clauses are derived via movement to Spec, CP of a Q(relative)‑P(article).

23 february
Vera Gribanova (Stanford)
Domains of identity in Russian TP ellipsis

This talk presents the results of an in-progress investigation into the explanatory source of non-isomorphism between antecedents and elided material in clausal ellipsis, across languages and configurations. Ellipsis non-isomorphism has typically been understood as the consequence either of a semantic identity condition (Merchant, 2001; Potsdam, 2007; Thoms, 2013), or of lexico-syntactic conditions whose definition permits some forms of non-isomorphism in a principled way (Thoms, 2015; Murphy, 2016; Ranero, 2021). Whatever the commitment about the formulation of the identity relation, the domain being compared with an antecedent for identity has usually taken to be coextensive with the ellipsis site itself. Recent findings—from Hebrew argument ellipsis of CPs (Landau, 2023), Spanish Wh-Topic-Remnant Elided Questions (Stigliano, 2022), and English sluicing (Rudin, 2019; Anand et al., To appear, 2021)—challenge this presupposition, suggesting instead that the domain under evaluation for identity may (or must) be a proper subset of the domain that is elided.

This set of moves opens new sets of questions and investigative directions, including the question of what the domain-based approaches to ellipsis identity might lead one to expect about crosslinguistic variation in ellipsis identity, and about the acquisition of such patterns. The talk explores three potential responses to such a state of affairs. First, there is the possibility that (i) the non-isomorphisms in e,g. Spanish, Hebrew, and English can still be derived from a theory that takes the domain of identity to be co-extensive with the ellipsis site itself, employing a looser isomorphism condition that permits certain mismatches. Second, there is the possibility that (ii) the domain evaluated for identity may be either coextensive with the elided constituent, or a proper subset of the elided constituent — defining a new axis of variation. Finally, there is the possibility that (iii) the domain of ellipsis identity is systematically smaller than the elided constituent.

The talk aims to discriminate among the positions in (i)-(iii) on the basis of broad predictions they make for ellipsis across languages and configurations. It presents novel, primarily corpus-based, evidence from three types of Russian TPE ellipsis — sluicing, stripping, and fragment answers — and compares those results against the predictions. It builds an argument toward a view of ellipsis identity that is rooted in a view like (ii) or (iii), and makes some initial attempts to discriminate among those two positions.

16 february
Stefan Keine (UCLA)
Silencing the PCC

Basque displays what is known as the Strong Person Case Constraint (PCC): an absolutive DP may generally not be 1st or 2nd person if it is c-commanded by a dative DP. We demonstrate that this restriction is surprisingly obviated under verbal ellipsis, even if this ellipsis does not affect the DPs whose cooccurrence is normally ruled out. We then explore the consequences of this generalization for accounts of the PCC. First, it indicates that the PCC arises from properties of the verbal agreement, not of the DP arguments. Second, a comprehensive account of the Basque PCC must be sensitive to both narrow-syntactic and PF properties (in particular whether or not the verb agreement is pronounced). We then develop an account of the Basque PCC based on Coon and Keine’s (2021) feature-gluttony proposal. On this account, the PCC results from an irresolvable conflict that arises in the morphological realization of a probe that has agreed with two DPs. We argue that this line of analysis offers a principled account of both the syntactic factors and the PF factors that condition the Basque PCC, in particular its interaction with verbal ellipsis. (This is joint work with Jon Ander Mendia)

15 february
Ethan Poole (UCLA)
Case and minimal compliance

Case depends on a limited set of syntactic configurations: head case, dependent case, and unmarked case. Why these configurations/types and not others? In this project, I explore the idea that the licit case configurations follow from the interplay between the operation Agree, a faithfulness constraint prohibiting tampering, and the Principle of Minimal Compliance.

9 february
Rodrigo Ranero (UCLA)
Against a What-You-See-Is-What-You-Get approach to Null Complement Anaphora: A Mayan perspective

Null Complement Anaphora (NCA) is the phenomenon in which the non-nominal complement of a verb is missing on the surface. Imagine a scenario where an invited speaker is arguing vociferously that all silent expressions are derived through witchcraft. An audience member can turn to another and utter either of the following: (1) Do you agree with what he’s saying?!, or simply (2) Do you agree?!

What is uncontroversial regarding NCA is that it patterns unlike other silent expressions along several dimensions. For example, Hankamer & Sag (1976) were the first to notice that NCA is unlike sluicing in allowing for pragmatic control: whereas NCA can be licensed via the context alone, (as in (2)), sluicing requires a syntactic antecedent, not a mere pragmatic one. Furthermore, Napoli (1983) and others noticed that NCA disallows sub-extraction, unlike other silent expressions such as VP-ellipsis. These contrasts have been taken to suggest that there is no complex syntactic structure underlying the “missing” element in NCA.

A healthy debate has thus arisen regarding what exactly the “missing” element is in NCA. Some argue that there is no missing element at all: What-You-See-Is-What-You-Get (WYSIWYG approaches; Shopen 1972, 1973, Napoli 1983, Grimshaw 1979; see Culicover & Jackendoff 2005, 2019). In other words, NCA verbs like ‘agree’ are akin to ‘eat’, which can vary in transitivity. Others argue that there is a null proform in NCA (Depiante 2001, 2019, Haynie 2010; see Hankamer & Sag 1976, Cinque 2004).

In this talk, we provide evidence from Chuj (Mayan) that NCA cannot be analyzed along WYSIWYG lines, supporting thus the view that a null proform underlies the construction. We argue that the universal syntax of NCA is revealed by the language’s morphosyntactic insistence to wear transitivity alternations on its verbal sleeve. We end by discussing why we should not interpret our results to suggest that the syntax of NCA should be parameterized across languages. (This is joint work with Justin Royer)

2 february
Ivy Sichel (UC Santa Cruz)
Parasitic gap licensing without movement

A leading empirical claim in the study of parasitic gaps (henceforth, PGs) is that they are licensed by an A-bar movement chain which stands in a particular structural relation to the PG (Engdahl 1983, and many others). Here I argue, based on the distribution of PG licensing by Hebrew RPs, that the requirement is actually weaker, and should be recast in terms of a licensing A-bar chain, which may or may not involve actual movement. I will show that a non-movement A-bar dependency culminating in a resumptive pronoun (RP) may also license a PG, as long as the structural condition on the relation between the licensing chain and the PG is met. This structural condition is identical to classical cases of PG licensing (Nissenbaum 2000) dubbed by Arregi & Murphy (2023) Parasitic Scope (following Barker 2007):

(1) Parasitic Scope for PG licensing: The PG-containing XP must take the scope of the binder of the licensing variable.

Crucially, I show that the variable may be realized as an A-bar Operator bound RP *which does not involve movement*. In the talk, I present differences and similarities in distribution between Hebrew PGs licensed by gaps and PGs licensed by non-movement RPs. The differences follow, I argue, from differences in the placement of the Operator heading the chain and determining the placement of the PG-containing XP (1). Whereas in movement chains launched from the complement of V, the immediate binder is the trace of movement left in specvP (Nissenbaum 2000), in A-bar binding chains which do not involve movement, the immediate binder is in specCP. Since the PG-containing phrase must take the scope of the licensor’s binder, this derives the core difference in PG-licensing: PGs contained within vP-level clausal adjuncts are licensed by a movement chain but not when the RP is not associated with movement (Sells 1984, and more recently Hewett 2023, based on much broader cross-linguistic coverage) – the binding Operator is too high. In other environments, where the Op binder does satisfy (1), these RPs do license PGs, showing that a movement derivation is not absolutely required.

26 january
No meeting. Have fun at NELS 54!

19 january
NELS practice talks

  1. Syntactic ergativity without inversion in Kalaallisut (Amy Rose Deal, Line Mikkelsen & Ellen Thrane)
  2. A novel way to diagnose phonological vs. suppletive allomorphy: Progressive STAMP morph formation in Lobi (Kang Franco Liu & Sansan Claude Hien)