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, 3:30-5:00pm


1303 Dwinelle Hall


Nico Baier

Katie Sardinha

previous semesters

spring 2014
fall 2013
spring 2013
fall 2012
spring 2012
fall 2011
spring 2011
fall 2010
spring 2010
fall 2009
spring 2009
fall 2008
spring 2008
fall 2007
spring 2007
fall 2006
spring 2006
fall 2005


University of California, Berkeley
Department of Linguistics


11 may
Yaron McNabb (Universiteit Utrecht):

Cross-categorial intensification: semblance or identity?

Recent analyses of cross-categorial intensifiers, such as really and definitely, take their flexible semantics and wide distribution as support for a uniform semantics, whereby they are universal quantifiers over contexts (McNabb, 2012; Beltrama & Bochnak, forthcoming). However, Bylinina & Sudo (forthcoming) have raised concerns regarding the overgeneration of such analyses. Moreover, while these analyses focus on crosslinguistic variation, a systematic comparison between (non-degree) intensifiers within a language is absent, and so additional observations regarding the contribution of these intensifiers to the management of propositions in the discourse are missed.

The goal of this talk is twofold: First, I propose a set of diagnostics I argue should be applied in any empirical investigation of intensification, illustrating it with several Hebrew intensifiers. These tests don’t only evaluate the semantic versatility of the intensifiers in question but also examine the intensification effect on the discourse status of the proposition. Second, I argue that in addition to quantifying over contexts, similar intensifiers affect the management of propositions, namely in settling Questions Under Discussion (Ginzburg 1996, Roberts 1996 and Büring 2003), keeping track of participants’ discourse commitments, and updating the common ground, couching my analysis in Farkas & Bruce’s (2009) definitions of discourse components. The empirical and analytical picture that emerges may question the viability of a uniform semantics but support the idea of shared semantic components with different effects on the discursive context.

8 may
Ryan Bochnak (UCB):

Optional tense? Evidence from Washo

As is by now well-known, the study of temporal reference cross-linguistically has revealed that many languages do not have tenses, and moreover that there are a variety of linguistic and contextual means that work to constrain the temporal reference of a clause. Susbequent research has viewed cross-linguistic variation in tense systems as a dichotomy between tensed and untensed languages. Nevertheless, some recent typological work (Plungian & van der Auwera 2006) has pointed to the existence of languages with apparently optional tenses. I take up this issue by investigating the semantic properties of the verbal suffix -uŋil in Washo, which is only compatible with past time reference, but is not obligatory in clauses with past temporal reference. Alongside clauses containing this morpheme, Washo also has morphologically tenseless finite clauses which can receive present, past or future temporal interpretations. That is, -uŋil is optional in the sense that its presence is not required for past time reference. I propose that -uŋil is the morphological exponent of a feature [PAST] that restricts the value of a temporal pronoun corresponding to the reference time of a clause. Furthermore, in morphologically tenseless clauses, tense features are simply absent, just like in many analyses of tenseless languages. The conclusion about cross-linguistic variation is that languages vary in terms of whether (interpretable) tense features are obligatory (English), absent (tenseless languages), or optional (Washo) in finite clauses. This analysis has consequences for analyses of apparently optional inflectional categories across languages, as well as for the nature of alternatives (e.g., Maximize Pressupposition, Heim 1991/Sauerland 2002), since the competition between tensed and untensed clauses does not result in restrictions on the use of tenseless clauses with past time reference in Washo.

1 may
James Collins (Stanford)

The mysterious case of disappearing indefiniteness: indefinite and definite patients in Tagalog

I discuss the semantics of Tagalog transitive patient NPs and the interaction of indefiniteness with verbal morphology and word order. In Tagalog verb-initial sentences, the verbal infix -um- is associated with non-specific transitive patients, while the verbal infix -in- is associated with definite transitive patients. However, in agent-initial sentences, the verbal infix -um- is obligatory and the infix -in- is ungrammatical via the morphosyntactic constraints of the Philippine-type symmetrical voice system (Schachter and Otanes 1984 a.o.). Although -um- is associated with indefinite patients in verb-initial sentences, -um- may be associated with either indefinite or definite patients in agent-initial sentences. The indefinite restriction on the patient of the -um-infixed verb disappears in agent-initial sentences.

Previous accounts of this phenomenon have been syntactic in nature (Rackowski 2002, Aldridge 2005) associating the syntactic position of a noun phrase with its (in)definiteness. However, the extant syntactic accounts rely on non-standard syntactic and semantic assumptions in order to link the syntactic position of the patient with the syntactic position of the agent. I suggest an alternative approach -- the paradigm falls out straightfowardly from neo-Gricean pragmatics.

Under my account, -in- forces a definite interpretation of the transitive patient, while -um- underdetermines whether the transitive patient is definite or indefinite. In a verb-initial sentence, the speaker may choose either affix. In interpreting the utterance of a sentence with the infix -um-, the hearer reasons that if the speaker had intended to convey a definite interpretation of the patient, the speaker could have chosen the unambiguously definite infix -in-. As she chose not to use -in-, the hearer reasons that the speaker must intend an indefinite interpretation of the patient. In an agent-initial sentence, the speaker only has -um- available to her (by Tagalog morphosyntactic constraints). Thus, the hearer can no longer assume that the speaker could have uttered the alternative sentence with -in-. Thus, the pragmatic reasoning doesn't go through. The conventional meaning of -um- is not pragmatically enriched, and is compatible with either an indefinite or definite interpretation.

24 april
Special joint meeting with Meaning Sciences: Zoltán Gendler Szabó (Yale)

Semantic Explanations

This paper is a defense of substantive explanations in semantics. I begin by offering a diagnosis of why the view that semantic theories are merely descriptive has been so widely accepted and I suggest that these grounds are not compelling. Then I argue that semantic explanations don’t have a uniform direction – upwards or downwards the syntactic tree. There is an explanatory division within the lexicon: the meanings of content words (nouns, verbs, adjectives, and adverbs) are semantically fundamental, while the meanings of function words (auxiliaries, connectives, copulas, derivational morphemes, determiners, expletives, prepositions, etc.) are derivative.

Link to paper

Note: This meeting will be held at a special place and time: Friday 4–5:30pm, 234 Moses Hall

17 april
Erik Zyman (UCSC)

Quantifier Float and the Driving Force for Movement: Evidence from Janitzio P’urhepecha

Many languages display the well-known quantifier float alternation:

(1a) All the walruses are painting.
(1b) The walruses are all painting.

Two main types of analyses of quantifier float have been proposed. On stranding analyses (Sportiche 1988, a.o.), (1b) underlyingly contains a nominal all the walruses, but the walruses moves out of this phrase, stranding all. On adverbial analyses (Torrego 1996, Benmamoun 1999, a.o.), all in (1b) is not a stranded adnominal quantifier but an adverbial element adjoined to some clausal projection.

I investigate this phenomenon in P’urhepecha, an isolate of central-western Mexico spoken mainly in the state of Michoacán, and specifically in the variety spoken on the island of Janitzio on Lake Pátzcuaro. In this language, the distribution of floating quantifiers tracks that of ordinary nominals extremely faithfully—a situation predicted by stranding analyses but unexpected on adverbial analyses. Some data involving Subject Condition effects initially seem to favor an adverbial analysis; but I argue that, despite this appearance, the facts of Janitzio P’urhepecha quantifier float strongly support a stranding analysis for this language, posing a challenge to “adverbial adjunction only” analyses of quantifier float.

Finally, having argued that quantifier float in Janitzio P’urhepecha comes about when a quantifier is stranded by its associate DP, I turn to the question of what drives DP movement generally in this language. I argue that the facts of DP movement in Janitzio P’urhepecha provide strong evidence for altruistic (target-driven) movement, and therefore constitute a challenge for analyses of movement (e.g., Bošković 2007) on which all movement is greedy.

10 april
Vera Gribanova (Stanford): The Verbal Identity Condition in ellipsis: Consequences for head movement

3 april
Kenny Baclawski (UCB): Topic and Focus Fronting in Cham

27 march
Spring Break

20 march
Maria Khachaturyan (UCB Anthropology)

Reflexives in Mano: less about syntax, more about pragmatics

In this paper I will analyse the syntax and semantics of reflexive markers in Mano (Mande < Niger-Congo). I will try to show that syntactic structure doesn't necessarily predetermine their distribution: other parameters, such as linear precedence and discursive prominence, are in play.

I will begin by the distribution of reflexive markers, I will study local and distant usages with subject controller. I will then show that the subject is not the only possible controller, and, more interestingly, that antecedents lower in the syntactic hierarchy than their anaphors can control reflexivity. I will show that linear precedence may play a more important role than syntactic structure.

I will then show examples of local reflexives used distantly. I will explain them by high discursive prominence of the referents, and show some typological parallels of the same phenomenon.

I will then turn to the distribution of simple and complex reflexive markers. Mano has a 3sg reflexive marker (ē); other persons and numbers don't have designated reflexive markers, but pronouns can also be used reflexively. Intensifier dìè is used to form complex reflexive markers (such as ē dìè). In fact, some verbs used in reflexive construction don't combine with simple markers, only with complex markers; some allow both. I will explain it by semantic particularities of the verbs. My functional explanation with be based on pragmatic factors (the speaker corrects listener's attention with the help of intensifier added to the structure of reflexive marker), in line with some typological literature on intensifiers.

13 march
Sophia Dandelet (UCB Philosophy):

A partial semantics for talking about partial desires

Sally eyes the last piece of apple pie. There is a part of her that wants to eat it, because it looks delicious. But there is another part of her that doesn’t want to eat it, since she believes that it would be rude to take the last slice of pie.

This paper is about sentences that report on desires like Sally’s. More precisely, it is about apparently contradictory desire reports — like “x wants p and x doesn’t want p,” or “x wants p and x wants q, but x doesn’t want both p and q” — and how we can make sense of them, from a philosophical and a linguistic perspective.

I begin by showing how the standard semantics for wants fails to make sense of these examples. Then, I develop an alternative analysis that makes the desired predictions for apparently contradictory desire reports. This semantics differs from its predecessors in two important respects. First, it assumes a situation-theoretic framework instead of the standard possible worlds framework. Second, it is sensitive to a parameter that allows us to set aside certain situations for the purposes of evaluating wants claims. I conclude by noting that my linguistic proposal points towards a philosophical understanding of desire on which wants may be held, so to speak, by parts of an agent — parts that take into account some of the questions and possibilities that the “whole” agent is concerned with, but ignore others.

6 march

27 february
Trio of Talks

Richard Rhodes (UCB) on Ojibwe bipartite verb structure and valence
Spencer Lamoureux (UCB): Multiple Copy Spell-Out as Constraint Interaction in Wandala (Central Chadic)
Nico Baier (UCB): An Overview of Cross-Linguistic Variation in Anti-Agreement

20 february
Syntax Circle Round Robin:

Everyone who wants to can have up to X number of minutes to discuss anything they want to related to syntax/semantics, such as an idea, a question, a problem, a controversy, or an interesting theory (where X = a number between 5 and 10, depending on how many people want to talk). Names will be drawn from a hat, so the order of turns is random. Share whatever's been on your mind!

13 february
No Meeting

6 february
No Meeting

30 january
No Meeting

23 january
No Meeting