Whamit!

Speaker: James Cooper Roberts (MIT)
Title: A Kalinian perspective on internal reduplication and its consequences
Time: Thursday, May 2nd, 5pm - 6pm
Location: 32-D769

Abstract: The true identity of infixes was a topic of much speculation in some corners of morphology. However, it was not until Kalin (2022) that we saw any hard evidence to inform an analysis. Considering cross-linguistic evidence from allomorphy, Kalin concludes that infixes are indeed (underlyingly) prefixes or suffixes. That is, infixes begin their lives at the edge of the root and are pushed to their pivot inside of the root. In her eyes, infixation is a finely timed process where exponent choice must precede infixation, which in turn must precede prosodification.

Kalin’s proposal is an exciting one, but there is one noteworthy gap in her proposal. In a Distributed Morphology framework, reduplication is accomplished by a special morpheme often glossed as RED. This morpheme is phonemically underspecified, instead consisting of a set of copying instructions which duplicate segments from a root. If we assume that internal reduplication is likewise a case of infixation, and RED enters the morphosyntax the same way any other morpheme does, when does this copying occur? Furthermore, is this order universal, or is there variation cross-linguistically? In a survey of genetically-diverse languages, Roberts (2023) finds that there is in fact cross-linguistic variation on when reduplication occurs relative to infixation. From this perspective, local internal reduplication can be thought of as a case where copying happens after infixation, and non-local internal reduplication is a case where copying happens before infixation.

However, proponents of Optimality Theory and related phonological frameworks may take issue with the assumptions of Roberts (2023) and Kalin (2022). For one, there is good reason to conclude that infixation is at least sometimes the result of phonological optimization rather than an arbitrary morphological process (e.g., McCarthy (2003)). Furthermore, reduplication can be thought of simply in terms of a correspondence between an output and itself. In a constraint-based framework, the facts of internal reduplication as they currently stand can be analyzed with independently-motivated constraints. Ergo, critics may rightfully wonder whether the theoretical machinery employed in Roberts (2023) (time of reduplication, “direction” of copying, etc.) is really necessary to account for the data.

In this presentation, I discuss these competing analyses in detail. I begin with an overview of Kalin (2022) and a detailed summary of Roberts (2023). I then discuss the analytical cost of such an approach to internal reduplication, and what we gain from it. This is followed by a summary of OT approaches to infixation and reduplication as both separate and combined phenomena, and I conclude with a discussion on the predictions of the two theories and how future work could inform the debate.

Speaker: Gianluca Porta (Ulster University)
Title: Causing extraction
Time: Tuesday, April 30th, 1pm - 2pm
Location: 32-D461

Abstract: A well-known distinction between arguments and adjuncts is that only the former allow extraction. A number of reasons for the island status of adjuncts have been proposed (e.g. Haung’s (1982) CED). Recent experimental studies challenged the categorical island status of adjuncts. In this talk I will present data that suggest that there can be extraction from them, and I will propose a theory to account for the data. I will show that an adjunct becomes transparent when there is a relationship of causation between the adjunct (the causing event) and the matrix clause (the caused event). I will apply this theory to two types of adjuncts: English temporal clauses and purpose clauses. While these clauses differ in a number of ways, they offer similar insights w.r.t. to what makes an adjunct transparent.

Speaker: Jon Rawski (MIT) and Zhouyi Sun (MIT)
Title: Tensor Product Representations of Phonological Constraints and Transformations
Time: Monday, April 29th, 5pm - 6:30pm
Location: 32-D831

Abstract: A crowning achievement of connectionist modeling in phonology embedded symbolic structures using tensors as an intermediary to neural computation, and used optimization over such structures to compute well-formedness (a la OT, HG, etc). However, there has been considerable difficulty restricting these models to match the upper computational bounds of phonology (regular languages and functions) since almost every constraint-interaction formalism computes supra-regular patterns with ease, and many are either Turing complete or uncomputable. We will discuss our recent attempt to circumvent this gap, by directly embedding both (sub)regular constraints and transformations into the tensor calculus used by constraint-interaction models. We will use finite model theory to characterize objects like strings, trees, graphs, and even input-output pairs as relational structures. Logical statements meeting certain criteria over these models define various classes of constraints and transformations. The semantics of such statements can be compiled into tensors, using multilinear maps as function application for evaluation. We show how this works for varieties of First-order and Monadic Second-Order definable constraints and transformations, and compare to previous work on correspondence constraints using model theory.

Speaker: Tanya Bondarenko (Harvard)
Title: TBA
Time: Thursday, May 2nd, 12:30pm - 2pm
Location: 32-D461

Abstract: TBA

This past weekend, April 26-28, the 27th Workshop on Structure and Constituency in Languages of the Americas (WSCLA 27) was hosted at the University of Toronto. Several current and former MIT linguists presented at the conference:

• Peter Grishin (Postdoc, PhD 2023): The Syntax and Semantics of Passamaquoddy-Wolastoqey modals
• Line Mikkelsen, Emily Clem, Michelle Yuan (PhD 2018), and Ellen Thrane (unaffiliated): Apparent Cross-Clausal Agreement with Obliques in Kalaallisut is Prolepsis
• James Crippen (McGill) and Jessica Coon (PhD 2010): Linguistics for Indigenous Language Study: A Course in Development
  Marta Donazzan, Hamida Demirdache (PhD 1991), Ana Lucia Müller, and Hongyuan Sun: On the Interpretation and Analysis of Covert Tense: Evidence from Karitiana

Speaker: Isabella Senturia (Yale)
Title: On the Spectra of Syntactic Structures
Time: Tuesday, April 23rd, 1pm - 2pm
Location: 32-D461

Abstract: This talk explores the application of spectral graph theory to the problem of characterizing linguistically significant classes of tree structures. As a case study, we focus on three classes of trees, binary, X-bar, and asymmetric c-command extensional, as well as different types of syntactic movement, and show that the spectral properties of different matrix representations of these classes of trees provide insight into the properties that characterize these classes. More generally, our goal is to provide another route to understanding the structure of natural language, one that does not come from extensive definitions and rules taken by extrapolating from the syntactic structure, but instead is extracted directly from computation on the syntactically-defined graphical structures. We also discuss implications of this work for generative capacity and for the Minimalist program.

Speaker: Logan Swanson (Stony Brook)
Title: Phonotactic Learning with Abductive Principles
Time: Monday, April 22nd, 5pm - 6:30pm
Location: 32-D831

Abstract: A foundational learning problem in phonology is phonotactics: discovering the constraints which govern what configurations of sounds are licit in a particular language. These constraints are generally formulated over some kind of substructure, which could be segmental, feature-based, or autosegmental. These substructures form a partial order, with simpler structures contained within more complex ones.

In this talk I will present the Bottom-Up Factor Inference Algorithm (BUFIA), an algorithm which leverages this underlying structure to learn surface-true phonotactic constraints from positive data (Chandlee et al., 2019). In this sense, BUFIA offers a general form for phonotactic learning, since it can operate over any choice of representation and use any abductive principle to decide when to add constraints. While Rawski (2021) examined several possible constraint selection principles, my work has shown additional kinds of abductive pressures that can guide the search, including execution of the search path itself in addition to other constraint selection criteria.

I will discuss how the BUFIA framework provides insight into the impact of these factors on learning, and give a demo of the software implementation.

Speaker: Yiyang Guo (University of Cambridge/Harvard University)
Title: Event counting, eventuality, and aspect
Time: Thursday, April 25th, 12:30pm - 2pm
Location: 32-D461

Abstract: Counting in the domain of events can exhibit a two-level nesting structure, i.e., the event level and the occasion level (Cusic 1981), as illustrated by the use of stacked time-adverbials in English (Andrews 1983; Ernst 1994; Cinque 1999) and two types of verbal classifiers in Mandarin Chinese (Deng 2013; Donazzan 2013; Zhang 2017). In Mandarin Chinese, event counting interacts with eventuality types and aspect. In this talk, I will provide two main observations:

(i) Countability of events hinges on eventuality types. Specifically, statives cannot be counted, activities allow counting at both the occasion level and the event level, while achievements and accomplishments can only be counted at the occasion level.

(ii) Counting expressions display difference in their compatibility with aspect. They are incompatible with the progressive marker -zhe, compatible with the perfective marker -le, but non-uniformly compatible with the experiential perfect marker -guo.

To account for the distribution of counting expressions, I will propose an atom-based analysis of event-counting (cf. Bach 1986; Krifka 1989; Landman 2006) under the framework of neo-Davidsonian event semantics (Parsons 1990; Carlson 1984; a.o.).

Speaker: Juan Cancel (MIT)
Title: Unexpected Syncretisms: A Look at the Nganasan Case and Subjective Agreement Paradigms
Time: Thursday, April 18th, 5pm - 6pm
Location: 32-D769

Abstract: Nganasan (Szeverényi, Várnai, and B. B. Wagner-Nagy 2002, B. Wagner-Nagy 2018) seems to have a near-exact, cross-paradigmatic syncretism between the NOM exponents of the Case Paradigm and 3rd Person exponents of the Subjective Agreement (The relevant syncretic exponents are highlighted):

Nganasan (Reduced) Case Paradigm
(B. Wagner-Nagy 2018, pg.191-193)

Nganasan Subjective Agreement Paradigm:
(B. Wagner-Nagy 2018, pg.229)

*In the grammar, ACC.DU and GEN.DU have a “ghost” consonant at the end.

I will argue for the following three things:

1) NOM.DU /-Kəj/ and 3rd.DU /-gəj/ are in fact the same exponent, meaning that we can argue that NUM is the morphosyntactic feature behind the syncretism and that there is no contextual allomorphy (Bonet & Harbour 2012) between nominal and verbal stems in the CONTEXT of NOM and 3SG.

2) The apparent syncretism between NOM.SG, ACC.SG and GEN.SG /-∅/ is superficial in nature since phonological phenomena such as Syllabic Gradation and Epenthesis clearly distinguish word-paradigms involving NOM.SG /-∅/ from word-paradigms involving ACC.SG and GEN.SG /-∅/.

3) The apparent syncretism between NOM.PL and GEN.PL /-ʔ/ is also superficial for similar reasons, but in light of the typology of epenthesis and vowel harmony (Finley 2008), GEN.PL /-ʔ/ is better understood as being /-Vʔ/, where V is a vowel that participates in vowel harmony.

Finally, one can see similar syncretisms in the Case and Subjective Agreement Paradigms of Tundra Nenets (Nikolaeva 2014, pg. 57, 59, 61, 78) and Forest Enets (Siegl 2013, pg. 121-124), suggesting that this cross-paradigmatic syncretism is more widespread in the Samoyedic language family.

Speaker: Janayna Carvalho (UFMG)
Title: Generic null impersonals in Brazilian Portuguese and structure removal
Time: Tuesday, April 16th, 1pm - 2pm
Location: 32-D461

Abstract: In this presentation, I explore some sentential properties of generic null impersonals in Brazilian Portuguese (BP). In particular, I look for an explanation for two of their properties: a) their restricted modal readings; b) their apparent ban on sequence of tense readings. I argue that these two disparate properties are due to the removal of part of the functional sequence of these clauses, due to a conflict between two probes – C and T – that could potentially value the external argument in these sentences.

Speaker: Sarah Payne (Stony Brook University)
Title: Marginal Sequences as a Window into Phonotactic Acquisition
Time: Thursday, April 18th, 12:30pm - 2pm
Location: 32-D461

Abstract: Most current theories of phonotactic learning (e.g., Hayes & Wilson 2008, Chandlee et al. 2019) assume a close relationship between attestation and licitness. Under such accounts, a sequence is licit only if its subcomponents (e.g., sequences of phones or feature bundles) are all attested in the input; illicit sequences are thus those that contain some unattested subcomponent. Under such theories, however, what is the status of marginal sequences (e.g., English ?[#sf])? Constraint-based views posit that marginal sequences are illicit but attested, making them an exceptional subclass of illicit sequences. However, marginal sequences pattern much more closely with licit sequences than illicit ones in terms of repairs in borrowings and in terms of production and perception errors, suggesting that they may instead be an exceptional subclass of licit forms. I argue for a theory of the phonotactic grammar in which attested sequences are divided into productive/licit ones and unproductive/marginal ones. I present a syllable-based computational learning model that learns a binary classification of attested forms into marginal or licit. When evaluated on English complex onsets, I show that this model matches well with human judgments, outperforming the model of Hayes & Wilson (2008) while accounting for the unique behavior of marginal sound sequences.

Speaker: Shrayana Haldar (MIT)
Title:Fixing Engdahl’s Type-Shifter and Heim’s Unary Which
Time: Wednesday, April 10th, 1pm - 2pm
Location: 32-D461

Abstract: Engdahl’s (1986) account of functional readings of sentences like (1) involved having the pronoun herself bound upstairs by a covert binder E, given in (2), while having a totally impoverished trace (i.e., just “t”, without any restrictor) with a complex index; that is, “tf(x) “. The E operator also does the job of shifting the restrictor to type ⟨ee, t⟩, without which the functional reading wouldn’t be possible. Heim (2019) pointed out that having the pronoun bound upstairs like this cannot derive ϕ-feature agreement between the pronoun and the antecedent quantifier, in this case, no girl, and binding theoretic effects like *Which picture of herself1 did no girl’s1 father submit?. Motivated by this reason, she proposed to have an LF like the one in (3), where which is unary (because it attaches directly and only to the question skeleton) and polymorphic (because it needs to be able to quantify over ⟨e, e⟩-type functions). Moreover, she proposed to have the whole restrictor picture of herself is in situ, getting syntactically bound by no girl, thereby avoiding the ϕ-featural and Binding Theoretic issues.

Speaker: Dóra Takács (MIT)
Title: Vowel-zero alternations in Hungarian
Time: Monday, April 8th, 5pm - 6:30pm
Location: 32-D831

Abstract: There are a closed class of about 500 stems in Hungarian that are subject to vowel-zero alternations if they are followed by a potentially vowel-initial suffix (Siptár & Törkenczy 2000). Early proposals (Vago 1980, Törkenczy 1992, Siptár & Törkenczy 2000, Abrusan 2005) mostly focused on whether this process is a result of epenthesis, metathesis or syncope . Some patterns among the consonants in these exceptional stems have been previously noted in the literature, but these observations were not integrated in the previous analyses. In this talk I show how stem-internal vowel-zero alternation in Hungarian interacts with voicing assimilation, affrication and gemination and use these interactions to support the claim that this process is in fact a result of syncope.

On Friday, 22 March 2024, James Cooper Roberts (first year student) presented his work on a computational investigation of the productivity of medials in Passamaquoddy at the Seventh Workshop on Computational Methods for Endangered Languages (ComputEL-7). This workshop was colocated with 18th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2024) in St. Julian’s, Malta.

Speaker: Stanislao Zompì (Universität Potsdam) and Zhouyi Sun (MIT)
Title: *ABA in Multidimensional Paradigms: A Harmonic Grammar-based account
Time: Monday, April 1st, 1pm - 2pm
Location: 32-D461

Abstract: In piece-based and realizational models of morphology, a fundamental question concerns the assignment and conditioning of phonological content to morphemes represented as feature bundles. This talk shows the limitations of Underspecification (Halle 1997; Bobaljik 2012) and Overspecification (Caha 2009; Starke 2009) approaches (henceforth US and OS) in deriving ABA patterns of contextual allomorphy in multidimensional paradigms (Christopoulos & Zompì 2022; Caha 2023). Our proposal essentially combines US and OS (cf. Ackema & Neeleman 2005; Wolf 2008; Müller 2020) and equates exponent choice to finding the ‘shortest path’ for each feature bundle to a legitimate exponent under a Harmonic Grammar framework. We further discuss the typology of ABA patterns of root suppletion and show that the proposed model makes appropriate restrictions on accidental homophony and derives the correct typology. Lastly, we suggest that our proposal yields output-driven maps in the sense of Tesar (2014), which boosts the learning of morphological paradigms.

Speaker: Fangning Ren (University of Cambridge)
Title: Approaching Mandarin wh-topicalization/focalization: D-linking effect
Time: Tuesday, April 2nd, 1pm - 2pm
Location: 32-D461

Abstract: This talk mainly provides two findings and the proposed analyses: (i) To form a content question, Mandarin ex-situ wh-nominals are invariably D-linked (Pesetsky 1987), whereas all wh-nominals can stay in situ regardless of their D-linking status; (ii) Mandarin wh-ex-situ creates both gapped dependencies and resumptive dependencies, with the former showing A’ diagnostics not the latter. Comparing three types of Mandarin wh-nominals: which-complexes, what-complexes and the what-simplex, I propose their varying degrees of D-linkability (in a morphosyntactic sense) are determined by their ability to realize the higher D head in a split-DP structure, which maps to the discourse (Roberts 2001, Guardiano 2012; Roberts 2017). In terms of the syntactic nature of the two dependencies created by Mandarin wh-nominal reordering, I argue that the resumptive one involves a base-generated wh-hanging-topic that introduces a null copular construction containing a reduced co-varying restrictive-relative head (Safir 1986, Kallulli 2012), whereas the gapped one involves wh-focus-movement. This is corroborated by the asymmetry they show in terms of parasitic-gap licensing, CNPC sensitivity, and the WCO effect.

Speaker: Yurika Aonuki (MIT)
Title: Minimum-standard predicates as resultatives and measure phrase interpretations
Time: Wednesday, April 3rd, 1pm - 2pm
Location: 32-D461

Abstract: In this practice talk for WCCFL, I will propose a compositional analysis of verbal predicates in Japanese that have been treated as minimum-standard gradable adjectives (GAs) (e.g., Kubota 2011; Sawada and Grano 2011). I demonstrate that the verbs in these predicates are inchoative-state verbs (Kiyota 2008) and argue that their GA-like behaviours provide novel empirical support for a state-based analysis of GAs (Wellwood 2015).

What have been treated as minimum-standard GAs, e.g., katamui-tei- ‘tilted’, allow absolute MPs (1). This property, in contrast with the lack of absolute MP readings with relative GAs (2), has played a major role in Japanese degree semantics.

(1) Poster-ga 5mm katamui-tei-ru poster-nom 5mm tilt-tei-npst ‘The poster is 5mm tilted.’ (2) Kono ki-wa 8m taka-i this tree-top 8m tall-npst ‘This tree is 8m taller.’ *‘..8m tall.’

However, these minimum GA-like predicates are morphologically Verb + aspectual marker -tei- (Oda 2008). I present the first compositional analysis of (1) in which the MP measures a result state.

Speaker: Heidi Duressi (MIT)
Title:An analysis of the Albanian verbal paradigm with with multiple exponence and paradigm contrast
Time: Monday, April 1st, 5pm - 6:30pm
Location: 32-D831

Abstract: Albanian verbs are typically analyzed to be broadly split into 2 categories: those with roots ending in vowels (Class 1) and those with roots ending in consonants (Class 2) (Newmark, Hubbard, & Prifti 1982; Camaj, 1984). There are certain segments (alternating between j and n) that are present in the indicative present forms of some verbs in Class 1, but not those of Class 2. There have been attempts at further characterizing these alternations (Trommer 2013) but there is no complete explanation of why they surface in the person/number combinations that they do. The goal of this talk is: 1) to present an analysis of this phenomenon, which combines ideas of multiple exponence, as suggested and used by (Matthews, 1974; Müller, 2006; Müller & Trommer, 2007), with paradigm contrast (Rebrus & Törkenczy, 2005; Hall, 2007) and 2) to explore the implications of this analysis for the general Albanian verbal paradigm.

Speaker: Noa Bassel (UMass)
Title: Complex anaphors
Time: Thursday, April 4th, 12:30pm - 2pm
Location: 32-D461

Abstract: Anaphors with a complex morphological structure are attested across many different languages and show recurring traits in meaning and distribution. Despite extensive research into these phenomena, there is no broadly-accepted definition of the primary grammatical function(s) of complex anaphors. This talk will evaluate various definitions from previous literature, with the goal of defining complex anaphors in a way that would explain their cross-linguistic properties: morphological complexity, binding effects, and homophony with focus-intensifiers (e.g., The queen HERSELF came to our party).

Speaker: Amanda Rysling (UC Santa Cruz)
Title: What it takes to comprehend (a) focus
Time: Friday, April 5th, 3:30pm - 5pm
Location: 32-141

Abstract: Over the past half-century, psycholinguistic studies of linguistic focus have found that comprehenders preferentially attend to focused material and process it more “deeply” or “effortfully” than non-focused material. But psycholinguists have investigated only a limited subset of focus constructions, and we have not come to an understanding of how costly focus is to process, what factors govern that cost, or why the language comprehension system behaves in the way that it does, and not others. In this talk, I discuss the problem for language comprehenders presented by the category of focus, and present evidence that focus processing is generally costly, but this cost can be attenuated by the presence of contrastive alternatives to a focus in the context before that upcoming focus. Evidence from the processing of second-occurrence foci demonstrates that comprehenders seem to work harder than our general models of sentence processing would posit that they should in comprehending given focused material. These findings add to our understanding of what it means to be good enough or efficient in language processing, delineating conditions under which comprehenders do (not) find apparently important material to be worth processing deeply or effortfully.

Speaker: Keely New (MIT)
Title: There’s no deletion in meN-deletion
Time: Tuesday, March 19th, 1pm - 2pm
Location: 32-D461

Abstract: In Indonesian/Malay, there is an optional verbal prefix meN- which is widely taken to be the subject voice marker since it correlates with subject voice SVO word order, and when it is present, only the subject may be A’-extracted. In object voice OSV word order, meN- is obligatorily absent, and only the object may be A’-extracted. The well-known “meN-deletion” generalisation is, therefore, that movement of the object over the verb in Indonesian/Malay results in deletion of meN (Saddy 1991, Fortin 2006, Aldridge 2008, Cole et al. 2008, Sato 2012, Georgi 2014 among others). Under such a perspective, the optional absence of meN- in subject voice is derived from a separate process from the obligatory absence of meN in object voice. Most analyses remain silent on the optionality of meN in subject voice. In this talk, I argue against a view where “meN-deletion” is triggered by movement of a DP across the verb. Drawing from data in Jakartan Indonesian, I propose that the choice between flavours of functional Voice/v head is one-to-one with the overt presence/absence of meN- prefix on the verb. In doing so, I argue that word order in Indonesian/Malay is but an epiphenomenal correlate of voice in the language.

Speaker: Lorenzo Pinton (MIT)
Title: Exclusive disjunction in bilateral logic: Hurford Disjunctions as evidence for split connectives in natural language
Time: Wednesday, March 20th, 1pm - 2pm
Location: 32-D461

Abstract: “Disjoint Hurford Disjunctions” (DHD; Amir Anvari, p.c.) are a novel class of examples of deviant disjunctions that resemble Standard Hurford Disjunctions (SHD; Hurford, 1974), but don’t present any classical entailment (or overlap) relation between the two disjuncts:

(1) SHD: # John lives in Paris, or he lives in France.

(after Hurford, 1974)

(2) DHD: # John lives in Paris and he’s married, or he lives in France and he’s single.

(Amir Anvari, p.c.))

In order to unify (1) and (2) under the ‘Hurford disjunction’ label, we have to make certain assumptions about how conjunction and disjunction work, and examples like (2) might therefore be particularly revealing of what type of logic is at play in language and reasoning. In this talk I will extend Bilateral State-based Modal Logic (BSML) from Aloni (2022) with an exclusive disjunction (ED). ED will be “classically” defined as supporting the disjunction of its disjuncts and rejecting their conjunction. Crucially, in BSML, supported disjuncts and rejected conjuncts are defined in a split way, and have therefore an ‘independent life’ from one another (for instance, to reject (A ⋀ B) you need to be able to reject A and B independently). I will claim that ED - implemented in a system that rules out zero models by default (i.e. BSML* from Aloni, 2022) - is a better predictor of assertability conditions of disjunction in natural language compared to both the classical logic and the standard BSML definitions of disjunction. First, equipping BSML* with ED provides a unified and direct semantic explanation to both standard Hurford Disjunctions and “Disjoint Hurford Disjunctions”. Second, ED yields the correct ‘uniqueness’ interpretation for sentences with multiple disjuncts, which has been a major challenge for past proposals of natural language disjunction as inherently exclusive. I will conclude the talk with possible challenges to the present system by showing that (i) conjunctions are not always split and we need a device to capture these cases (possibly along the lines of subject matter, from Truthmaker semantics (Fine, 2017)) and that (ii) assuming ED as the standard meaning for disjunction actually clashes with some results in Aloni (2022) (wide scope free choice) and Degano et al. (2023) (absence of exclusivity in production tasks). Time permitting, I will briefly sketch in-progress solutions to solve the conflicts.

Speaker: Yeong-Joon Kim (MIT)
Title: Overapplication opacity as a consequence of phonetic faithfulness
Time: Monday, March 18th, 5pm - 6:30pm
Location: 32-D831

Abstract: This study contributes to the understanding of opacity by identifying substantive restrictions on counterbleeding interactions and proposing a novel analysis tied to these typological generalizations. A typological survey of counterbleeding-on-environment instances reveals an asymmetry in the types of opaque processes involved, with assimilation and consonant-induced vowel processes being the most common. A novel account of phonological opacity is suggested to deal with this asymmetrical distribution of the opacified processes. The basic idea for explaining this observed asymmetry is that most opaque interactions have a functional rationale, that of preserving phonetic properties of lexical entries (e.g., Flemming 2008). The approach can also account for overapplication opacity in feeding interactions, such as self-destructive feeding in Japanese, which is problematic for classical Optimality Theory.

Speaker: Ksenia Ershova (MIT) and Nikita Bezrukov (UPenn)
Title: Moving away from antilocality: A defense of very local movement
Time: Thursday, March 21st, 12:30pm - 2pm
Location: 32-D461

Abstract: Most syntactic research assumes that syntactic dependencies are subject to locality constraints: agreement and movement cannot cross certain elements (bounding nodes, phase boundaries, elements which bear the same features, etc.). A growing body of work also argues that there are antilocality constraints which impose a lower bound on syntactic dependencies – movement, and perhaps Agree, must cross a certain type of boundary. Focusing primarily on Spec-to-Spec Antilocality (Erlewine 2016 et seq.), which states that movement from Spec,XP “must cross a maximal projection other than XP”, we argue that such constraints are unlikely to be part of Universal Grammar. The argument is three-fold. First, we briefly discuss the trajectory of antilocality theories and argue that they were originally proposed as responses to theoretical idiosyncrasies which don’t extend beyond the frameworks that they are couched in. Second, we focus on one of the most discussed empirical motivations for antilocality theories – constraints on subject extraction – and suggest that this data can be analyzed in other ways, while antilocality approaches fall short in explaining the patterns. And finally, we present a case study of possessor relativization in West Circassian which demonstrates that Spec-to-Spec Antilocality makes the wrong empirical prediction: very local movement exists.

Speaker: Shota Momma (UMass Amherst)
Title: A theory of structure building in speaking
Time: Tuesday, March 12th, 1pm - 2pm
Location: 32-D461

Abstract: Partly due to our free will, our ability to produce sentences is notoriously hard to study. Existing theories of sentence production are not very good at capturing how speakers assemble structurally complex sentences that involve syntactically interesting phenomena. In this talk, I attempt to fill this gap by developing a theory of sentence production that integrates well with theories of syntax. This model aims to capture the production of sentences involving various syntactic phenomena, including raising, control, and wh-movement dependencies, and suggests a close parallelism between locality domain for syntactic computations and planning units in sentence production. The proposed theory makes counter-intuitive predictions about the time-course of sentence formulation as well as one of the most well-studied phenomena in sentence production: structural priming. I present experimental evidence confirming those predictions.

Speaker: Caroline Heycock (University of Edinburgh) & Elise Newman (MIT)
Title: When to revisit? investigating (un)ambiguity in temporal clauses
Time: Wednesday, March 13th, 1pm - 2pm
Location: 32-D461

Abstract: Ever since the seminal work of Geis in the 1970s, it has been known that temporal adverbial clauses in English (and a number of other languages) show the same kind of ambiguities as also observed in wh-questions and (even more relevantly) relative clauses, so that a sentence like “She arrived just when I predicted she would arrive” can mean either that her arrival coincided with my act of predicting, or with the time given in the prediction. This follows fairly directly from an analysis of such temporal adverbial clauses as a type of free relative clause denoting a definite description of a time, which can be formed syntactically by A’-movement of a temporal operator, or whatever analysis is given to such long-distance dependencies. Here we’d like to report on joint work in progress (Caroline Heycock, Elise Newman, Rob Truswell) where we investigating cases of temporal clauses where long-distance movement seems to be either excluded or at the least heavily disfavoured, and exploring the possibility that the data motivate a distinction between temporal clauses that are descriptions of time intervals and those that are descriptions of events.

Speaker: Levi Driscoll
Title: Ludika: Playing with Feet
Time: Monday, March 11th, 5pm - 6:30pm
Location: 32-D831

Abstract: We present a description and formal analysis of a novel English-based ludling called Ludika. Starting from clear cases with direct cues to foot structure, we argue that meaningless bits are affixed to the right edge of surface English feet in Ludika. We then use Ludika as a diagnostic of foot structure in more ambiguous contexts such as initial stressless light syllables (po‘tato). Evidence from Ludika aligns with Kiparsky’s (1979) characterization of these initial stressless syllables as unstressed degenerate feet, rather than syllables adjoined to feet in a recursive structure (Jensen 2000, Davis & Cho 2003, Kager 2012) or stray syllables (Ito & Mester 2009). We also leverage phrasal data to illustrate that function words in phrases behave just like feet elsewhere in Ludika, suggesting that they are neither stray syllables in phonological phrases (Selkirk 1996) nor syllables adjoined to content words as part of a recursive prosodic word structure.

Speaker: Rajesh Bhatt (UMass Amherst)
Title: A blocking effect in Hindi-Urdu
Time: Friday, March 15th, 3:30pm - 5pm
Location: 32-141

Abstract: In this talk, I will present a blocking effect in Hindi-Urdu that could be characterized as Poser blocking, namely a word winning over a phrase. This blocking arises in the context of deverbal adjectives and its analysis has implications for the representation of adjectives as a syntactic category and the structure of verbal derivations in Hindi-Urdu.

Many languages have specialized morphology that can verbalize adjectives. In English, for example, -en can combine with a number of adjectives (e.g. flatten, redden). Hindi-Urdu lacks verbalizing morphology of this sort but it has another device that allows for productive verbalization of adjectives. This very productive device involves the light verbs ho ‘be’ and kar ‘do’. There is, however, a restriction on the application of this very productive device: the class of deverbal adjectives cannot be verbalized via the light verb strategy. I propose that we can see verbalization as being blocked by the existence of equivalent lexical verbs; hence the characterization in terms of Poser blocking.

However I do not derive this effect by appealing to the notion of Poser blocking as a primitive. Instead I explore a derivational system where the `blocked’ cases are simply not derived; instead the derivational process delivers the corresponding lexical verbs. The derivational system I set up has a surprising feature: a semantically contentless categorizing head needs to be treated as a last resort element that is inserted countercyclically.

Phonology Circle will meet on March 4, 5-6.30pm, in the 8th floor conference room for a logistical meeting. Be there or be square!

On February 22, 2024, our fifth-year grad student Ido Benbaji-Elhadad gave an invited talk at the interdisciplinary colloquia series organized by the Linguistics Department at Tel Aviv University. 

Title: Specific-opaque readings and the temporal interpretation of noun phrases

Abstract: Szabó (2010,2011) argues that in addition to their de dicto, de re and third readings, DPs in intensional contexts can have a fourth, specific-opaque reading in which their determiner scopes above an intensional operator while their restrictor is nevertheless interpreted opaquely, in the intensional environment created by the operator. Szabó provides examples of specific-opaque readings relative to the three main intensional operators, namely, attitudes, modals and tense, as evidence that natural language makes available a general, unrestricted mechanism to derive such “split” readings for DPs in intensional environments. We focus on specific-opaque readings relative to modals and tense and show that neither supports that conclusion; i.e., that both are restricted in meaningful ways and should be derived without a general mechanism. The discussion serves to highlight the ways in which modal and temporal operators differ with respect to the availability of the specific-opaque reading for DPs in the intensional environments that they create, providing new evidence that worlds and times differ in the kind of mechanisms that introduce them to the semantic composition.

Speaker: Heidi Harley (University of Arizona)
Title: Object drop, intention, and event individuation (joint work with Raffaella Folli)
Time: Friday, March 8th, 3:30pm - 5pm
Location: 32-141

Abstract: We introduce a pattern of interaction between object drop and the animacy of the external argument in English, particularly with verbs of contact. Object drop produces an agentivity/animacy constraint on the newly intransitive verb, which doesn’t exist in the transitive equivalent. Compare John pushed the car to the curb and The glacier pushed the boulder to the sea. The former, but not the latter, allows for object drop: At the signal, John pushed! vs #In winter, the glacier pushed.

Our analysis follows Martí 2015, according to which object drop instantiates noun incorporation of a null indefinite N. As with all noun incorporation, this tends to yield a ‘conventionalized’ denotation for the VP—compare Michael lifted the glass with Michael lifted.

In both object drop and (overt) noun incorporation contexts, the denotation of the object nominal is nonreferential—no object participant in the verbal event is introduced. Instead, the object nominal is integrated into the predicate denotation via some flavor of predicate restriction or predicate modification (Chung & Ladusaw 2004, e.g.) We contend that the absence of an event participant creates a challenge for the application of the predicate by the speaker. A predicate which has been subject to object drop or noun incorporation, then, lacking an event participant, does not pick out the same class of events that the transitive verbal predicate does. Instead, we contend, intentionality is recruited to derive the sortal content of the verbal predicate. Martin 2015 showed that an agent’s intention by itself can constitute aa so-called “indicative property”, providing enough individuating properties to license the application of a change-of-state predicate in nonculminating contexts even when the change of state encoded by the verb has not yet begun to occur. Without an intentional agent, she shows, such ‘Zero-CoS’ readings of telic predicates are unavailable, since no indicative properties which can identify the ongoing event as an instance of the predicate exist until the CoS actually begins to occur. Intentions, then, can identify events even when ‘normal’ event identification criteria, such as the configuration of participants or the existence of a CoS, do not apply. This, we propose, is why the subjects of such predicates have to be (intentional) Agents—it’s because without intention, the indicative properties necessary for application of the predicate cannot be identified. This also lets us understand why object drop contexts so often produce such specialized ‘conventionalized community activity’ readings—it’s a subcase of the wider case where simply intending to execute an action of a particular type yields enough indicative properties to license predication. We term this effect the ‘Goal Oriented Condition’ on object drop. This condition on event individuation in turn constrains the types of subject argument that are compatible with object drop contexts, since external arguments must be ‘teleologically capable’ of executing the event type denoted by the predicate. In the case of intentionally individuated event types, the only kind of external argument that has the relevant teleological properties are animate Agents.

Over the weekend, the linguistics department at UMass Amherst hosted the annual ECO-5 workshop. ECO-5 is a workshop for graduate students of five departments on the East Coast (UMass, Harvard, UConn, UMD, MIT) to present work in progress to each other. This year, MIT was represented by two second-year graduate students, Zachary Feldcamp and Bergül Soykan:

• Zachary Feldcamp: Evidence for low base-generation of PP in locative inversion
• Bergül Soykan: Plural Marked Interrogatives in Turkish

Our MITILI student Rani Ukhengching Marma was recently featured on 7000.org for International Mother Language Day. The feature highlights the important work that Ukheng is doing to preserve and protect her native language Marma. Some of this work is ongoing right now as part of her MA thesis work, but Ukheng also shares about her commitment to language and cultural preservation beyond the dissertation. Read more here: https://www.7000.org/post/our-global-community

Speaker: Christopher Legerme (MIT)
Title: Movement Dependencies and Existential HAVE constructions in Haitian Creole
Time: Tuesday, February 27th, 1pm - 2pm
Location: 32-D461

Abstract: For this upcoming Syntax Square, I will be presenting content from two papers which are relevant for an unresolved puzzle of Haitian Creole syntax: 1. Takahashi and Gracanin-Yuksek’s (2008) Morphosyntax of Movement Dependencies in Haitian Creole and 2. Myler’s (2018) Complex copula systems as suppletive allomorphy. Here’s the puzzle. The Haitian Creole (HC) sentence kimoun ki te genyen is ambiguous.

1. (a) Ki moun *(ki) te genyen?
      WH person COMP PST win.
“Who won?” (lit. Which person that won?)

(b) Ki moun *(ki) te genyen?
WH person COMP PST have
“Who was there?” (lit. Which person that had?)

The surfacing of the complementizer ki after a wh-moved phrase is claimed to be symptomatic of subject extraction (Koopman 1982, DeGraff 2001, Takahashi and Gracanin-Yuksek 2008). Therefore, it’s puzzling that ki obligatorily surfaces in (1b) given how the suppletive allomorphy approach (Myler 2018) deals with the syntax of existential “copula” constructions like the following:

2. (a) Te genyen yon moun
      PST have a person
        “There was a person.” (lit. had a person)

(b) Yon moun *(ki) sanble (ki/∅/*ke) te genyen
INDF person COMP seem COMP PST have
“There seemed to be a person.” (lit. A person that seems that had)

Note that in (2b), the second ki is optional. To my knowledge, these facts of HC are unexplained in the literature. Related facts are the obligatory presence of ki with moved thematic arguments of certain (transitive?) verbs like jwe “play” or anbrase “hug”.

   3. (a) (Se) gita ki t ap jwe nan pak la
it.is guitar COMP PST PROG play in park the
“GUITAR was playing in the park.” (lit. (It is) GUITAR that was play in the park)

(b) (Se) ti moun yo ki t ap anbrase
It.is small person PL COMP PST PROG hug “THE KIDS were hugging” (lit. (It is) THE KIDS that was hug)

HC apparently doesn’t show subject inversion (DeGraff 1992: 48) nor English-style passive morphology (Deprez 1992: 208). Still some verbs inflect depending on whether its external argument is agentive or thematic, that is, there are verbs in HC that alternate between a transitive and intransitive form (DeGraff 2001: 75).

4. (a) Mwen fè/*fèt kabann lan rapid-rapid maten an
1SG make bed the fast-fast morning the “I made the bed very quickly this morning.” (DeGraff 2001: 75)

5. (b) Kabann lan *fè/fèt rapid-rapid maten an
bed the made fast-fast morning the
“The bed was made very quickly this morning.” (DeGraff 2001: 75)

I do not have a solution to these puzzles, but I will present the analyses of complementizer ki and existential HAVE sentences of Takahashi and Gracanin-Yuksek (2008) and Myler (2018), respectively, to show how exactly a sentence such as (1b) is puzzling given the theories advanced in those papers, and I will speculate that the correct solution could come from how one might handle the null expletive there and complementizer agreement in the language.

Speaker: Omri Doron (MIT)
Title: Disjunctive inferences and presupposition projection
Time: Wednesday, February 28th, 1pm - 2pm
Location: 32-D461

Abstract: Multiplicity and Homogeneity (demonstrated in 1 and 2 below) are cases of a truth value gap: the negated sentences in (1b) and (2b) are stronger than what we would get by applying logical negation to the sentences in (1a) and (2a). Evidence gathered in recent years indicates that these truth value gaps are the result of disjunctive presuppositions triggered by the (in)definite plural NPs. Both sentences in (1) presuppose that either Jack saw more than one horse or no horse, and both sentences in (2) presuppose that either Mary read all of the books or none of them. However, it is still unclear how these presuppositions come about. In this talk, I propose that they are the result of presuppositional exhaustification applied at the NP level, and show that their disjunctive nature can be cashed out as a consequence of the properties of presupposition projection from the environments in which they are generated.

(1) a. Jack saw horses.
Inference: Jack saw more than one horse.
b. Jack didn’t see horses.
Inference: Jack saw no horses.

(2) a. Mary read the book.
Inference: Mary read all of the books.
b. Mary didn’t read the books.
Inference: Mary read none of the books.

Speaker: Christopher Legerme (MIT)
Title: Verbal (Non-)Apocope in Haitian Creole
Time: Thursday, February 22nd, 5pm - 6pm
Location: 32-D769

Abstract: Verbal apocope is widely attested across French-lexified creole (FLC) languages and it involves the alternation between a short and long form of the verb in certain syntactic contexts (Seuren 1990, Henri and Abeillé 2008).

(1)  Mo fin mâze (*mâz)
      1.SG finish eat
       “I have eaten.”

(2) Mo fin mâz diri la (*mâze)
      1.SG finish eat rice DET
     “I have eaten the rice.” (Mauritian Creole, Seuren 1990)

(3) Konbyen dan Tonton Bouki genyen (*gen)
      how-many tooth Uncle Bouki has
      “How many teeth does Uncle Bouki have?”

(4) Tonton Bouki gen 32 dan l (*genyen)
       Uncle Bouki has 32 tooth 3.SG
      “Uncle Bouki has (all of) his 32 teeth.” (Haitian Creole, DeGraff 2001)

In (1) and (3), the long form of the verb is obligatory, and in (2) and (4) the short form is obligatory. The morphosyntactic or morphophonological conditions influencing verb shortening (or lengthening) are still a mystery and vary across FLCs, with linguists citing factors such as the (non-)presence of a VP complement or semantic agent, as well as other syntactic, prosodic, or semantic considerations depending on the language. Haitian Creole (HC) stands out as the most systematic in the realization of its verbs and most verbal roots of HC in fact must surface with the verbalizing affix -e, such as in manje “to eat” (√Manj-+-e) and genyen “to win” (√Gen+-e, with added phonological processes deriving from gen /gɛ̃/ the final surface form genyen /gɛ̃jɛ̃/). This verbalizing affix may also be -i for a certain class of roots, for example, fini “to finish” (√Fin+-i) or mouri “to die” (√Mour+-i). HC is also the least constrained in the domain of verbal apocope because, aside from (4) (which I will account for), there isn’t much evidence that HC short forms are ever obligatory to the exclusion of the long form, unlike in the apocope patterns of other FLCs where both the exclusively short and the exclusively long contexts are widespread. However, HC verbs that permit apocope (e.g., gad(e) “look”, gen(yen) “have”, fin(i) “finish”, vin(i) “come”) do show evidence that the long form is required in some contexts. For example, one generalization (attested in other FLCs, too) is that the short form cannot occur when the complement of the verb is absent. For this talk, I will demonstrate that the pattern of HC short/long alternation is actually that of a verbal non-apocope, where the shortening of the verb by deletion of the verbal suffix -e is barred by PF constraints rather than there being a licensing of the short form by a combination of various semantic or syntactic considerations. Crucially, this phenomenon of verbal non-apocope in HC is related to stress assignment at the interface of syntax and phonology and a constraint against apocope when the verbalizing affix -e has primary stress. Finally, I will extend my analysis to a famous puzzle of HC linguistics concerning the pattern of the ∅/ye “be” copula alternation. 

Speaker: Yurika Aonuki (MIT)
Title: Degree constructions in Gitksan
Time: Wednesday, February 21st, 1pm - 2pm
Location: 32-D461

Abstract: I will present my ongoing work on degree constructions in Gitksan (Ts’imshianic; see Rigsby 1986 and Bicevskis et al. 2017 for previous documentation). Based on the availability of comparative and superlative interpretations of positive forms as well as the obligatory differential readings of co-occurring measure phrases in positive constructions, I argue that gradable adjectives take a differential degree as an argument (1), which builds on Oda’s (2008) analysis of Japanese (see also Beck et al. 2004).

(1) [[‘tall’]] = λd. λx. λd. Height(x) ≥ d + d

An alternative account assuming a covert comparative or superlative opera- tor is ruled out by additional data showing that the comparative and superlative readings of positive constructions are unavailable with minimum-standard predicates. In addition, what appears to be a standard phrase in comparatives only indirectly determines the standard degree (see Hohaus 2015). Finally, as my analysis predicts, constructions with relative adjectives always involve comparison; those that do not involve such comparison, such as degree questions and absolute measurements, require a nominalizer, which I analyze as saturating the standard degree argument of a gradable adjective with a zero degree.

Speaker: Giovanni Roversi (MIT)
Title: Workshopping φ-marking in Äiwoo
Time: Tuesday, February 13th, 1pm - 2pm
Location: 32-D461

Abstract: Äiwoo verbs usually carry markers indexing the φ-features of the subject (agreement? Clitics/reduced pronouns? We’ll talk about it). Beyond this one basic fact, not much is neat or clear. These markers show different patterns in different voices, in terms of the morphological exponents, their position (prefixes vs suffixes), their ability to simultaneously co-occur with an overt subject DP, and more. Moreover, there are also cases where these markers index the object’s φ-features instead of the subject’s. And of course, a theory that correctly derives the behavior of the φ-markers is only a good one inasmuch as it can fit well into a general theory of Äiwoo syntax (word order facts, the morphosyntax of voice, patterns of Ā-extraction, etc.). I have the data, now please help me out!

Speaker: Amir Anvari (MIT)
Title: Revisiting E-Type Theory
Time: Thursday, February 15th, 12:30pm - 2pm
Location: 32-D461

Abstract: A implementation of E-Type Theory faces two serious problems, the problem of uniqueness and the problem of formal link. In this talk, I will discuss a simple implementation of E-Type Theory that addresses the problem of uniqueness with minimal auxiliary assumptions. The predictions of the resulting theory for a number of cases, including anaphora to wh-indefinites, will be examined in detail. To address the problem of formal link, I will flesh out a rule that I will refer to as Antecedent Conversion based on the proposals in Parsons 1978 and Heim 1990. AC prevents the theory from over-generation. It also makes a prediction regarding the underlying form of pronouns that are anaphoric to nested indefinites which in turn leads to interesting predictions involving crossover phenomena and i-within-i effects which I will discuss as well.

Welcome back everybody from winter break, whether it was restful or productive! Here is a round-up of what people have been up to: 

• The 100th LSA meeting was held in NYC January 4-7! 
  • Anton Kukhto, along with Alexander Piperski (Stockholm University, Sweden), gave a talk Stress clash avoidance in Russian comparatives: A corpus study. 
  • Megan Gotowski and Forrest Davis gave a talk titled What Language Models Can Tell Us About Learning Adjectives.
  • Keely New gave a talk titled Voice and the variable position of auxiliaries in colloquial Jakartan Indonesian.
  • Agnes Bi gave a talk titled Tonal identification in whispered speech. 
  • Omri Doron and Jad Wedbe gave a poster titled Scalar implicatures are sensitive to constraints on presupposition accommodation. 
  • Dora Takacs was a panelist at the panel Securing an internship, contract work and part-time jobs during your degree program: A practical guide for linguistics students & faculty advisors. The panel was organized by the LSA Linguistics Beyond Academia Special Interest Group.
  • Maya Honda was a speaker at the Session Collaborative Efforts in Linguistics: Partnerships Between and Among Secondary and Higher Education Institutions.
  • Johanna Alstott gave a talk titled A first semantics for at first and at last. 
  • Chelsea Tang, along with Sansan C Hien (University of California, Berkeley), gave a poster titled STAMP morphs in Lobi: Morphological and Typological Implications. 
  • Katie Martin gave a poster titled A stereotype-based semantics for slurs.
  • Suzanne Flynn, along with Chie Nakamura (Waseda University) and Katsuo Tamaoka (Nagoya University), gave a poster titled L1 vs. L2: Persistence of processing cost due to differences in relative clause configuration.
  • Christopher Legerme gave a poster titled Why Haitian Creole ye-tracing is non-verbal predicate resumption. 
  • Lorenzo Pinton and Janek Guerrini gave a poster titled Numerous-like predicates in bare plural generics. 
  • Jon Rawski organized a Special Session with Scott Nelson and Jeff Heinz (Stony Brook), titled Formal Language Theory in Morphology and Phonology.
    
    
• Jon Rawski gave two invited colloquium talks in Paris! They were: Learning (Sub)Regular Transformations at the Automata Seminar, Institut de Recherche en Informatique Fondamentale (IRIF), and Abductive Inference of Phonotactic Constraints at the Ateliers de Phonologie, CNRS.
  
  
• Jon Rawski and co-authors had their article The Problem-Ladenness of Theory accepted in the journal Computational Brain and Behavior! Congratulations. 
  
  
• Eunsun Jou had her article Honorification as Agree in Korean and beyond published in Glossa! Congratulations! 
  
  
• Hadas Kotek begins a 3-year term as LSA Executive Committee member-at-large.
  
  
• NELS54 was a great success, held January 26-27 at Stata. Many thanks to the organizers (Keely, Agnes, Giovanni, Zhouyi, Stan, Athulya, David), student helpers, and especially the staff, who made all of it possible! Pictures below. Discerning readers will also see that we had David Pesetsky’s birthday. Presentations were given by: 
  
  
  • Jad Wedbe, Covert reciprocals (talk)
  • Johanna Alstott, Before and after decomposing first and last (talk)
  • Yeong-Joon Kim, Phonetic faithfulness in counterfeeding opacity (talk)
  • Peter Grishin & Anton Kukhto, Infixing Outward (poster)
  • Adèle Hénot-Mortier , The French demonstrative paradigm: structurally transparent but semantically intricate (poster)
  • Haoming Li, Dou and plural universal quantification in Mandarin Chinese (poster)
    
    
• Norvin Richards, Peter Grishin, Elise Newman, Cora Leasure, and Cooper Roberts embarked on a Passamaquoddy trip to Maine! Pictures below. 

Speaker: Nicholas Rolle
Title: “Phonological locality and constraints on exponent shape”
Time: Friday, Feb 9th, 3:30pm - 5:00pm
Location: 32-141

Abstract:
The focus of this talk is exponence – the mapping of syntactic representation (e.g. features, nodes, small trees) to phonological representation (e.g. segments, tones, etc.) via stored X↔Y pairings. What are the restrictions on the contents of Y (the exponent) in such pairings? While Optimality Theory established principles curtailing restrictions on underlying forms (e.g. “Richness of the Base”), approaching this issue from the syntax-phonology mapping reveals one robust constraint: all components of an exponent must be local, either in a contiguous string on a single phonological tier, or connected via an association line across tiers. To support this thesis, we examine two types of ‘bipartite morphemes’: circumfixes of the German ge-…-t type, and grammatical tone involving distinct segmental and tonal components. While bipartite morphemes superficially contradict our constraint, based on their morphological patterning we show that the multiple components always constitute separate exponents.

We are pleased to announce a series of five lectures by Prof. Matilde Marcolli. The lectures will be based on three recent papers by Matilde co-authored with Bob Berwick and Noam Chomsky. The talks will be hybrid. Contact Amir Anvari for any questions. 

Title: Mathematical Structure of Syntactic Merge: An Algebraic Model for Generative Linguistics 

Abstract: This series of lectures is based on work in collaboration with Noam Chomsky and Robert Berwick. The main goal is to present a mathematical formulation of Chomsky’s theory of Merge and the Strong Minimalist Thesis, and show how various aspects of the theory fall naturally into place in terms of the algebraic structure. We will discuss how one can think, in this light, about Externalization, about the difference between older forms of Minimalism and the new SMT, and about the interface between syntax and semantics.

Papers: [https://arxiv.org/abs/2311.06189], [https://arxiv.org/abs/2306.10270], [https://arxiv.org/abs/2305.18278]

Times and Locations:

• Monday 11^th of December: Room 32-D461, Time 12-1:30pm 
• Tuesday 12^th of December: Room 32-D831, Time 1-2:30pm
• Wednesday 13^th of December: Room 32-D461, Time 10-11:30am
• Thursday 14^th of December: Room 32-D461, Time 1-2:30pm
• Friday 15^th of December: Room 32-D461, Time 1-2:30pm

Please note the differences, both in time and location, of the lectures on different days.

Speaker: Johanna Alstott (MIT)
Title: Trying and failing to count in dense intervals
Time: Wednesday, December 13th, 1pm - 2pm
Location: 32-D461

Abstract: In this LSA practice talk, I offer a semantic analysis of a puzzling restriction on the distribution of ordinal numbers in English: while the temporal adverbials “at first” and “at last” are felicitous, putting any other ordinal in this environment is degraded (#at second, #at sixth). I know of no previous literature that discusses “at first”/”at last” or the unacceptability of #at second, #at third, etc. My analysis of “at first” and “at last” builds on the notion that assertions are relativized to a salient time interval, known in the literature as topic time (Klein 1994). On my semantics, “at first” and “at last” further relativize an assertion to a salient subinterval of the topic time that shares an infimum (first point) or supremum (last point) with it. On the assumption that time-intervals are dense, the infelicity of #at second, #at third, etc. follows from this semantics. Since “at first” and “at last” invoke the infimum and supremum of a time-interval (respectively) on my semantics, #at second will attempt to invoke the second (i.e. second earliest) point of a time-interval. Invoking the infimum or supremum of a (closed) dense interval is coherent, but invoking the second earliest point (the point closer to the infimum than any other) is not. My analysis makes interesting predictions about the interaction of “at first”/”at last” with present tense and with frame adverbials, and it paves the way for an account of why ordinals are forbidden in related “at”-modifiers with superlatives (e.g. at most vs. #at second most).

Speaker: Ksenia Ershova (MIT)
Title: What’s in a (polysynthetic) phase: Dynamic domains, spellout and locality
Time: Thursday, December 7th, 12:30pm - 2pm
Location: 32-D461

Abstract: INTRODUCTION This talk demonstrates how languages with complex morphology (=polysynthetic languages) can help tease apart rules which apply purely in the syntax from interface conditions which determine how linguistic structure is pronounced. The two types of rules nontrivially interact with each other, making it difficult to determine the division of labor in languages like English, which don’t make use of complex morphology. The results of the study highlight the importance of understudied, typologically diverse languages for our understanding of human linguistic competence.

ABSTRACT Since Chomsky (2000), constraints on locality have been defined in terms of phasehood. One explanation for the opaqueness of phases for syntactic operations has been to treat them as potential goals and, correspondingly, interveners for Agree with elements inside the phase (Rackowski and Richards 2005, etc.). On the other side, a broad consensus has been to connect phasehood to the timing of spell-out: syntactic structure is sent to PF cyclically, and a constituent which has been spelled out is no longer visible for higher syntactic derivations (Uriagereka 1999, Chomsky 2001, 2008, etc.). Building on this intuition, a number of approaches have identified phases as salient domains for defining prosodic rules (e.g. Newell 2008, Dobashi 2013).

Based on data from West Circassian, I argue for an integrated theory of phasehood which combines both approaches. West Circassian provides evidence for the existence of two partially overlapping, but independent notions of syntactic domain: (i) spellout domains which are relevant for defining rules of syntax-to-prosody mapping, and (ii) locality domains, which serve as interveners for Agree. The two types of domains display different properties: prosodic domains are spelled out wholesale, together with the head and the edge of the constituent, while locality domains are equidistant to higher probes with their edge, allowing for successive-cyclic movement out of them. Both types of domains are dynamic (cf. den Dikken 2007; Gallego 2010; Bošković 2014), but in different ways and under distinct conditions. Prosodic domains are defined by the boundaries of the extended projection: for example, an NP behaves as a prosodic constituent even in the absence of D. The opaqueness of locality domains, on the other hand, is relativized to a given probe – if the probe has independently agreed with the head of a locality domain, the corresponding domain no longer displays phasehood properties in relation to that probe.

As a polysynthetic language, West Circassian provides a unique window into the dichotomy between the two types of syntactic domains due to the parameters which derive its complex morphology. In the domain of syntax-to-prosody mapping, polysynthetic languages allow for systematic mismatches between syntactic and prosodic phrasing (e.g. Compton and Pittman 2010; Barrie and Mathieu 2016) – certain phrases correspond to phonological words and are thus identifiable as prosodic domains. The sensitivity of locality domains to Agree is observable due to obligatory agreement between heads of the same extended projection, which, among other things, drives concatenation of verbal heads to form complex wordforms – in certain well-defined configurations, this agreement renders locality domains transparent for probing, thus confirming Agree-based approaches to phasehood.

References Barrie, M. & E. Mathieu. 2016. Noun incorporation and phrasal movement. Natural Language and Linguistic Theory 34: 1–51. Bošković, Željko. 2014. Now I’m a phase, now I’m not a phase: On the variability of phases with extraction and ellipsis. Linguistic Inquiry 45.1.27–89. Chomsky, Noam. 2000. Minimalist inquiries: The framework. In R. Martin, D. Michaels & J. Uriagereka (eds.), Step by step: Essays on minimalist syntax in honor of Howard Lasnik, 89–156. Cambridge, MA: MIT Press. Chomsky, Noam. 2001. Derivation by phase. In M. Kenstowicz (ed.), Ken Hale: A life in language, 1–52. Cambridge, MA: MIT Press. Chomsky, N. 2008. On phases. In Foundational issues in linguistic theory, eds. R. Freidin, C. P. Otero & M. L. Zubizarreta, 133–166. MIT Press. Compton, R. & C. Pittman. 2010. Word-formation by phase in inuit. Lingua 120: 2167–2192. den Dikken, Marcel. 2007. Phase extension: Contours of a theory of the role of head movement in phrasal extraction. Theoretical Linguistics 33.1–41. Dobashi, Yoshihito. 2013. Autonomy of prosody and prosodic domain formation: A derivational approach. Linguistic Analysis 38. 331–355. Gallego, Ángel J. 2010. Phase theory. John Benjamins. Newell, Heather. 2008. Aspects of the morphology and phonology of phases. Doctoral dissertation, McGill University. Rackowski, A. & N. Richards. 2005. Phase edge and extraction: A Tagalog case study. Linguistic Inquiry 36 (4): 565–599. Uriagereka, Juan. 1999. Multiple spell-out. In S. D. Epstein & N. Hornstein (eds.), Working minimalism. Cambridge, MA: MIT Press.

Speaker: Benjamin Spector (Institut Jean Nicod (CNRS-ENS-EHESS))
Title: Reasoning with Quantifiers, Lewisian Imaging and the Confirmation Paradox
Time: Friday, December 8th, 3:30pm - 5pm
Location: 32-141

Abstract: From a normative point of view, the conclusions we can draw from a sentence of the form No A is B are the same as the ones we can draw from No B is A, because No is a symmetric determiner, and therefore these two sentence types are logically equivalent. I will present experimental evidence (based on joint work with Vincent Mouly) showing that people do not in fact reason in the same way with No A is B and with No B is A. In particular, people’s estimate of the number of As after processing No A is B is higher than after processing No B is A (and vice-versa for B). I will argue that this instantiates a more general property of restrictors: we tend not to revise our beliefs about the size of the restrictor set even when receiving information that would in fact warrant such a revision. I will argue that this effect can be explained if we make the following hypotheses:

- Belief update does not (always) correspond to probabilistic conditionalization, but can also proceed by Imaging, as defined by David Lewis (1976). In a nutshell, when revising our beliefs with a proposition S, our posterior degree of confidence in a certain proposition T corresponds to our prior degree of confidence in the conditional ‘If S, T’ (using Stalnaker’s semantics for conditionals), rather than to the conditional probability P(T|S).
- Restrictors tend to serve as anchors when we engage in conditional reasoning: when considering the different ways in which a quantified sentence could be true, we mentally keep constant the restrictor set. I will relate this both to the possibility of de re readings for restrictors and to recent experimental results about verification strategies for quantified statements (Knowlton, Pietroski, Williams et al. 2023).

I will show that how these findings and the proposed theory can shed light on the confirmation paradox (see also Rinard 2014): given a statement S of the form ‘All As are Bs’, people are more prone to think that an observation of an object that has both properties A and B ‘confirms’ S than they are to think that observing an object that is both not-B and not-A confirms S, and I will discuss, time permitting, further experimental results (based on joint work with Nicolas Poisson) pertaining to the confirmation paradox.

Selected References:

- Knowlton, T., Pietroski, P, Williams, A., Halberda, J & Lidz, J. (2023), Psycholinguistic evidence for restricted quantification. Nat Lang Semantics 31, 219–251. https://doi.org/10.1007/s11050-023-09209-w
- Lewis, D., (1976), Probabilities of Conditionals and Conditional Probabilities, Philosophical Review, 85(3): 297–315. doi:10.2307/2184045
- Rinard, S. (2014), A New Bayesian Solution to the Paradox of the Ravens, Philosophy of Science 81 (1):81-100 (2014)

Speaker: Sigrid Beck (University of Tübingen)
Title: The emergence of a Generalized Quantifier: English ‘every’
Time: Friday, Dec 1, 3:30pm – 5pm
Location: 32-141

Abstract: The talk investigates the diachronic development of universal quantifiers in English. The etymological source of Present Day English ‘every’, Old English ‘aelc’ is analysed as an indeterminate pronoun. It participates in an alternative semantic system of quantification. I show how during Middle English, this system was reduced and finally lost. A reanalysis as a Generalized Quantifier becomes possible towards the end of the Middle English period. The compositional semantic change is analysed. I argue that interpretive stability at a propositional level is a crucial condition on such semantic change.

Speaker: Noa Bassel (UMass Amherst)
Title: No choice: Anaphoric dependencies in the prepositional domain
Time: Tuesday, November 28th, 1pm - 2pm
Location: 32-D461

Abstract: Pronominal elements have long been understood as cues for the length of linguistic dependencies. This is based on a regular morphology that targets pronouns in local dependencies and leaves long-distance pronouns unmarked, leading to the well-known complementarity between complex anaphors and simple pronouns.

The category of prepositions creates a gap in this respect, as it introduces positions in which the division of labor between anaphors and pronouns is inconsistent. In these environments, pronouns and anaphors may have the same reference (e.g., Max rolled the carpet over him/himself).

This has motivated claims that that PPs enable free choice between the anaphor and the pronoun, which speakers employ to convey nuanced semantic details and attitudes. Accordingly, P anaphors have so far seemed as an unreliable syntactic diagnostics and were largely ignored by theories of P syntax.

I will argue that P anaphors are restricted to local dependencies as in any other environment, and that the apparent freedom in pronoun choice follows from a robust structural ambiguity in the domain of prepositions.

Speaker: Magdalena Lohninger (MIT)
Title: Cross-clausal A-dependencies: A composite probe approach
Time: Thursday, November 30th, 12:30pm - 2pm
Location: 32-D461

Abstract: Hyperraising, i.e. A-movement out of a CP complement clause, posits a puzzle for generally assumed constraints like the Ban on Improper Movement or the Phase Impenetrability Condition and has been discussed widely in the recent years. In this talk I present a typological examination of Hyperraising as well as Hyper-ECM across unrelated languages, covering them under the umbrella term cross-clausal A-dependencies (CCA). I start by investigating empirical properties of CCA such as that i) the A-dependency stems from the matrix predicate, ii) the DP undergoing CCA is base-generated in the embedded clause and moves through the embedded left edge (SpecCP) and iii) the embedded CP often resembles a regular phase. In the course of this typological investigation of CCA, I conclude that the ability to hyperraise is not a parametric option but rather, that CCA falls into five structurally different configurations across languages (including Prolepsis). These five resemble each other on the surface but underlyingly differ in four relevant properties: i) selectional properties of the matrix predicate, ii) movement or high base-generation of the DP undergoing CCA, iii) A-Minimality differences and iv) interpretational restrictions (s.a. topic requirements). Based on these four parameters, I present a methodological tool to disentangle CCA constructions across different languages as well as within single ones. Further, I suggest a composite A’/A probe analysis for a subclass of CCA constructions, namely such involving movement out of the embeddded clause. The analysis is couched in ongoing research on the probing mechanism of A’/A probes and I discuss different implementational options such as Feature Gluttony (Coon & Keine 2021, Coon, Baier & Levin 2021), Interaction and Satisfaction (Deal 2015, 2022) and Contingent Probes (Branan & Erlewine 2020, Branan 2021). Last, I delve into (last resort) independent probing options for composite A’/A probes and what they might tell us about CCA co-occuring with regular long-distance A’-movement.

Christopher Legerme, Cora Lesure, and Lorenzo Pinton represented MIT Linguistics at Fall Splash on November 18-19. Splash is an MIT student-run educational enrichment program for 9th-12th graders that draws a thousand students to MIT each November for a weekend of classes on all sorts of topics. Christopher and Lorenzo designed and taught “The Mathematical Foundations of Language” and Cora designed and taught two sections of “What’s in a word? An Introduction to Morphology”. Maya Honda observed their classes and attests that all three did a fantastic job of sharing their knowledge and their passion for linguistics with 60 students. Many students stayed after class to ask questions and seek advice about studying linguistics. Kudos to Chris, Cora, and Lorenzo!

Speaker: Christopher Legerme (MIT)
Title: Pronominal Copulas and Defective T(ense) in Haitian Creole
Time: Tuesday, November 21st, 1pm - 2pm
Location: 32-D461

Hi, I will be presenting some of my ongoing work on the morphosyntax of copular clauses in Haitian Creole (HC), focusing on the puzzling distribution the word se (glosses ranging from “FOC”, “COP”, or “SE”), which behaves like a pronominal copula though has resisted a unifying account of its multifunctional properties and sparked much debate among linguists on a range of syntactic, morphological, and semantic issues (DeGraff 1992, 1995, Kihm 1993, 2018, Déprez & Vinet 1997, Déprez 2003). The gist of the puzzle is as follows: 

Puzzle 1: Why does se occur adjacent negation or tense/aspect “markers” such as te “PST” (let’s call them N(eg)T(esne)A(aspect) auxiliaries, cf. DeGraff 1992: 29-36) in some contexts but not others? For example, exclusively se can be used for the present-tense matrix copular clause (1a), and it cannot surface with NTA auxiliaries such as te (nor with non-property-denoting lexical predicates). Notwithstanding, in se-clefts (“it-clefts”) se appears to be in the position where we expect the subject (before the NTA auxiliary).

(1) a. Saika *(se) yon doktè 

         Saika SE     a     doctor 

         Saika is a doctor

b.  Saika  (*se) te         (*se)    yon doktè 

     Saika     SE  PAST       SE  a      doctor 

     Saika was a doctor 

c. Se   te     medikaman mwen (ke)  Saika (te)   preskri 

    SE  PST medicine      my       that Saika PST prescribe 

    It was my medicine that Saika prescribed 

Puzzle 2: While se is generally obligatory before DPs (1a), why isn’t this so for clauses with wh-moved DP subjects, especially when the subject is a wh-word (DeGraff 1992: 122-123; Déprez & Vinet 1997: 221)? 

(2) a. Se Saika_i    ki          t_i  (se)   yon   doktè 

          SE  Saika  COMP       SE   a       doctor 

          It is Saika who is a doctor 

        b.  kimoun    ki           (??se)  yon doktè? 

              who        COMP       SE   a     doctor

              Who is a doctor?

Speaker: Shrayana Haldar (MIT)
Title: Modal Debris: Threefold Ambiguities between Permission, Weak Necessity, and Strong Necessity in Bengali
Time: Wednesday, November 22nd, 1pm - 2pm
Location: 32-D461

Abstract: The usual way to characterize the Bengali copular modal hoy is to say it’s the strong necessity modal of the language. I will show that hoy is ambiguous between strong necessity and weak necessity in upward-entailing contexts, permission and strong necessity under negation, and permission, weak, and strong necessity in polar questions. Following Staniszewski (2022) I will show why a scopal account, varying the attachment heights of negation and the modal, doesn’t work, using diagnostics involving the presuppositions of no longer and only. The account I will propose instead will involve pursuing the lines pursued by Staniszewski (2022), who derived weak necessity from strong necessity by (i) putting an existential quantifier over ordering source sequences on the strong necessity modal, (ii) exhaustifying this structure with existential force into an meaning of weak universal force by innocently including relevant subdomain alternatives and pruning the irrelevant ones (Bar-Lev 2018). The threefold ambiguity in polar questions will be accounted for, again following Staniszewski (2022), by making a silent even operator interact with the scope of whether (Guerzoni 2004). Time permitting, I will discuss some questions it raises about restrictions on forming Katzirian (Katzir 2007, Fox and Katzir 2011) deletion alternatives, a constraint I have been thinking about for the past few weeks, how I think it works better than Meyer’s (2013) restriction on deleting covert material, and also does away with the need of the Atomicity+ constraint developed in Trinh (2018).

Congrats to our MITILI student Ukhengcheng Marma, who was awarded the Priscilla King Gray (PKG) social impact fellowship for IAP 2024! 

Speaker: Zachary Satoshi Feldcamp (MIT)
Title: Paradoxical A-movement locality in locative inversion
Time: Tuesday, November 14th, 1pm - 2pm
Location: 32-D461

Abstract: Locative inversion has a locality paradox. It involves movement of PP over the would-be DP subject, and is restricted in non-finite clauses, like topics (Stowell 1981). Yet inversion only occurs when DP is in an exceptional rightward position (Doggett 2004), and otherwise adheres to A-movement locality conditions. In particular, I show that locative inversion cannot be formed via subextraction from DP, unlike PP topicalization.

I present an analysis that allows XP A-bar movement to exhibit A-movement locality. First, finite T has a disjunctive A-/A-bar probe. This allows XP with A-bar features to move to spec,TP, but only if there is no closer DP. We gain a new understanding of Barss’ generalization: reconstruction of DP into an XP predicate that undergoes A-bar remnant movement is impossible because DP could not have raised out of XP, XP being a closer goal to the disjunctive probe on T than DP.

Second, locality is based not on c-command, but on D-precedence: G D-precedes K iff G precedes K or G irreflexively dominates K. This allows XP with phi- or A-bar features to move to spec,TP over a c-commanding DP only when DP is linearized to the right of the position from which XP moves. I show that a key prediction is borne out: in passive constructions, rightward movement of the DP internal argument demotes DP and promotes the PP dative argument.

Speaker: Lorenzo Pinton (MIT)
Title: Two puzzles for gather-like and numerous-like predicates. Anti-restrictiveness, maximality, and triviality in plurals
Time: Wednesday, November 15th, 1pm - 2pm
Location: 32-D461

Abstract: Among collective predicates, gather (and alike) and numerous (and alike) have been noticed to give rise to different patterns when plural quantifiers are involved (Kroch, 1974; Dowty et al., 1987; Champollion, 2010; Kuhn, 2020). In the first part of the talk, I present an analysis for a new asymmetry between the two predicates in restrictive relative clauses noticed by Martin Hackl (p.c.):

(1) I only talked to the students that gathered.

(2) #I only talked to the students that were numerous.

I tie the different patterns to a difference in restrictiveness, showing that restrictive modification of pluralized predicates by numerous is blocked whenever a maximality requirement (like that of the definite article the) applies on top because of triviality (possibly in the style of Gajewski (2002). That is because numerous is permutation invariant (PI) with respect to members of sums, and when a PI predicate modifies a pluralized predicate (like *students), the maximal sums obtained before and after modification are identical. Crucially, modifying students with a collective predicate like gather makes restriction by numerous available (Hackl, p.c.):

(3a) I only talked to the gathered students that we numerous.

This is predicted, under the assumption that for collective predicates we can delay pluralization until after combination with numerous, in order to avoid triviality, giving rise to the following LF for (3a):

(3b) Jack only talked to [the [* [ [ [gathered][*[students]] ] […numerous] ] ]]

In the second part of the talk, I present joint work with Janet Guerrini on another previously unobserved asymmetry between gather-like and numerous-like predicates, this time involving bare plurals. Whereas the former can give rise to generic readings, the latter cannot:

(4) Gathered students are loud. (Generally, when students are gathered, they are loud)

(5) Numerous students are loud. (#Generally, when students are numerous, they are loud)

We present evidence for kind-based approaches' to bare plural generics (Chierchia, 1998; a.o.) over what we callambiguity approaches’ (Diesing, 1992; a.o.), showing that the latter overgenerate, predicting unobserved generic readings when numerous-like predicates modify bare plurals in (5). We claim that on kind-based approaches, such generic readings are directly ruled out because of a maximality operator which leads to triviality, following the analysis proposed in the first part of the talk. Evidence from ellipsis will show that GEN needs indeed to be polymorphic (Chierchia et al., 1995), but strongly constrained to bind the lowest possible type, given parallelism. I conclude the talk with crucial counterexamples that might undermine the analyses proposed, and for which I currently don’t have a full-fledged solution.

Speaker: Magdalena Lohninger and Yiannis Katochoritis (MIT)
Title: It’s getting out of control! – Control, the A/Ā-distinction and the subjecthood of pivots in Austronesian languages
Time: Tuesday, November 7th, 1pm - 2pm
Location: 32-D461

Abstract: Control has long been employed as diagnostic of subjecthood: in an active embedded clause, the controllee should invariably correspond to the agent qua external argument. Yet, the very notion of subjecthood breaks down in Austronesian-type voice systems where one argument is promoted to “pivot” and triggers a particular morphological form on the verb without altering its valency: subject properties are distributed between the agent irrespective of pivothood, and the pivot irrespective of θ-role, though not in a uniform manner (Schachter 1976; Guilfoyle et al. 1992).

The status of Austronesian voice-marking has been subsumed under three syntactic analyses: (i) pivots are established in vP/VoiceP via a process of object shift and subsequent Ā-Agree reflecting structural case, θ-role or extraction site; (ii) voice-marking reflects absolutive case assignment in a (split-)ergative system; (iii) pivots are (hanging or internal) topics.

We examine this debate on the basis of control structures, where Austronesian languages are divided into two types: type A, in which control targets the pivot (e.g., Malagasy, Acehnese), versus type B, in which control targets the external argument irrespective of voice-marking (e.g., Tagalog, Madurese). The picture gets murkier through the existence of embedded voice restructuring, default uses of Agent Voice morphology, backwards and crossed control configurations, and restrictions on matrix-embedded Voice matching.

At the same time, Austronesian pivots differ with respect to the definiteness/ specificity restriction and A/Ā-related properties such as anaphor binding. Particularly the latter forms two types of languages: type C, in which pivots reconstruct (e.g., Malagasy, Tagalog), versus type D, in which promotion to pivot feeds binding (e.g., Acehnese).

Given that types A & B of control do not fully overlap with types C & D of binding, we delve into the properties of pivots on a language-specific basis, and suggest that their A/Ā-nature is neither uniform nor absolute, but spreads over a continuum: some pivots are more A-like elements, some are more Ā-elements, and some are mixed, each determined by the type of features involved in the derivation of voice-marking. Such differences might stem from a diachronic transition from Ā-to-A syntax of voice marking (see also Chen & Patrianto 2023), each possibly correlating with several aspects of the clausal structure, such as word order, position of the non-pivot agent and allomorphy.

We conclude that the notion of “pivot” might be a structural epiphenomenon and subject to variation, in which case a uniform syntactic analysis of voice marking should be untenable.

Speaker: Jad Wehbe (MIT)
Title: Covert Reciprocals
Time: Wednesday, November 8th, 1pm - 2pm
Location: 32-D461

Abstract: Predicates like “date” and “hug” participate in alternations between seemingly 1-place variants in (1a) and (2a) and 2-place variants in (1b) and (2b). Analyses of these alternations can be grouped into syntactic analyses (e.g. Hackl, 2002) and lexical analyses (e.g. Winter, 2019). On the syntactic analysis, these predicates are always 2-place predicates, and the LF for (1a) and (2a) involves a covert reciprocal in object position. On the other hand, the lexical analysis assumes that the predicates date and hug in (1a) and (2a) are collective 1-place predicates, and that the 2-place versions are derived from the 1-place variants in the lexicon. A major challenge for the syntactic account comes from cases when the covert and overt reciprocal are not truth-conditionally equivalent, as in (3) (Winter, 2019). In this talk, I propose a version of the syntactic account that addresses this challenge. I argue that the truth-conditional differences are due to distributivity (“each”) having to take lowest possible scope when “each other” is ellided. I provide evidence for this account and against a lexical one from the behavior of these predicates in downward-entailing environments and in the Lebanese Arabic double subject construction. (1) a. Jane and Mary dated. b. Jane dated Mary. (2) a. Jane and Mary hugged. b. Jane hugged Mary. (3) Context: Jane hugged Mary while she was sleeping, and then Mary hugged Jane later while Jane was sleeping. a. # Jane and Mary hugged. b. Jane and Mary hugged each other.

Speaker: Vina Tsakali
Title: Interpretation of complex- and simple-Or in adult and child Greek
Time: Thursday, November 9th, 12:30pm - 2pm
Location: 32-D461

Abstract:

Disjunctive particles have been argued to differ cross- and intra-linguistically as to whether they are obligatorily exclusive or not (Szabolcsi 2001, Aloni 2016). The current project examines the differences between the simple (i=Or) and the complex forms (i-i and ite-ite=Either-Or) of the disjunctive operators in Greek.

The results of four online questionnaires on adult Greek speakers showed no significant effect of the disjunction type suggesting that Greek adult speakers do not perceive one of the two types as obligatorily exclusive.

The development of the disjunction in Greek tested via a forced choice picture matching test provides evidence that children interpret both disjunction types inclusively (and probably conjunctively) until the age of 7.

Speaker: Omri Doron (MIT)
Title: Reduplication and number in Colloquial Jakartan Indonesian
Time: Wednesday, November 1st, 1pm - 2pm
Location: 32-D461

Abstract: NPs in Colloquial Jakartan Indonesian (CJI) do not generally bear any overt number morphology, and are traditionally assumed to be number-neutral semantically. For example, (1) is judged to be true both when there is a single difficult question in the homework, and when there are multiple ones.

(1) PR-nya ada soal yang susah
homework exists question that difficult
“The homework contains (a) difficult question(s)”

One strategy that CJI uses to convey multiplicity is reduplication of the noun. (2) is judged to be true if there are multiple difficult questions in the homework, but false if there is only one. This multiplicity inference does not arise in DE environments - (3) is only judged as true if there are no difficult questions. This raises the question of what is the semantics of reduplication.

(2) PR-nya ada soal-soal yang susah
homework exists question-question that difficult
“The homework contains difficult questions”

(3) PR-nya nggak ada soal-soal yang susah
homework NEG exists question-question that difficult
“The homework does not contain difficult questions”

Dalrymple and Mofu (2012) analyze CJI reduplicated nouns as denoting a “relatively large” plural individuals, putting the weight of the analysis on the threshold for being relatively large in different linguistic contexts. As opposed to them, I argue that the semantics of reduplication is essentially identical to that of plural-marked nouns in languages like English. In my analysis, the difference between CJI and English lies in the alternative relation between different forms. This may allow us some insight into the morphology of number marking and reduplication across languages.

Speaker: Anastasia Tsilia (MIT) & Zhuoye Zhao (NYU)
Title: What the incompatibility of ‘then’ with the present teaches us about perspectives in tense
Time: Thursday, November 2nd, 12:30pm - 2pm
Location: 32-D461

Abstract: This talk focuses on the then-present puzzle, namely the observation that the present tense is incompatible with the temporal adverbial ‘then’ (Ogihara & Sharvit 2012; Vostrikova 2018; Tsilia 2021). This is attested not only in root clauses (e.g. *John is then not feeling well), but also in embedded clauses (e.g. *John thought that Mary is pregnant then) across languages such as Russian, Modern Hebrew, Modern Greek and Japanese where the embedded present can ‘shift’ to overlap not with the time of utterance, but with the time of the embedding attitude/speech. We propose to account for this generalisation by assuming i) that the present tense and ‘then’ are both sensitive to a temporal perspective shared by all expressions within a minimal clausal domain, and ii) the present and ‘then’ carry contradicting perspectival presuppositions. The perspective is modelled closely after the context as an interpretation parameter; we will briefly discuss their connections and differences.

On this basis, we further investigate the cross-linguistic behaviour of (embedded) clauses where ‘then’ co-occurs with past tenses. A particularly interesting case concerns the so-called deleted past tense, observed in languages such as English and Modern Greek, whose past tense feature is apparently uninterpreted when embedded under a higher past tense. The deleted past tense has long been thought to be semantically indistinguishable from the shifted present tense, but only the deleted past is compatible with ‘then’. The difference can be easily accommodated within the current analysis by assuming that unlike non-deleted tenses (including the shifted present), deleted tenses are not perspective-sensitive.

Speaker: Yoonjung Kang (UToronto)
Title: Speech Rate Accommodation under Sound Change in progress: Case Studies from Daejeon Korean
Time: Friday, November 3rd, 3:30pm – 5pm
Location: 32-141

Abstract: Variation in speech rate is a common characteristic of speech and serves as a significant source of synchronic variation and potential sound change. During fast speech, segments tend to shorten compared to normal or slow speech, potentially blurring the distinctions between long and short segments. In this presentation, I will discuss the results of two recent perception studies conducted on Daejeon Korean. One study focuses on non-low back vowels, while the other explores stop laryngeal contrasts. These studies aim to shed light on how listeners adapt to changes in speech rate when ongoing sound changes potentially neutralize the duration-based cues related to the target contrasts.

This last Saturday (10/28) was the warmest weekend before spring, and a group set out to Boston Hill Farm for apple and pumpkin picking! See pictures below. 

Our postdoc Peter Grishin gave two talks at the 55th Algonquian Conference this past weekend (Oct 20–Oct 22, 2023), hosted by the University of Alberta in Edmonton (https://algonquian-conference.org/pac50/):

Peter Grishin (MIT): Passamaquoddy-Wolastoqey modals
Peter Grishin (MIT) & Will Oxford (University of Manitoba): When central suffixes agree with peripheral participants

Check the slides out here and here! 

Speaker: Elise Newman (MIT)
Title: When wh-phrases are their own interveners
Time: Tuesday, October 24th, 1pm - 2pm
Location: 32-D461

Abstract: Much work on syntactic locality has shown that processes like wh-movement are subject to several kinds of locality restrictions. In addition to being sensitive to intervening wh-phrases, wh-movement must proceed successive cyclically through various points in the clause, and in some cases/languages, may not cross intervening arguments (see e.g. Branan and Erlewine (2022) for a recent overview). Sensitivity to intervening arguments is known to be quite fine-grained: according to insights from Keenan & Comrie (1977) and others, languages might differ with respect to what kinds of arguments count as interveners for a wh-element, and might also treat arguments vs. adjuncts differently.

In this talk, I propose that all of these locality restrictions and their various levels of granularity are interconnected. More specifically, I suggest that they reduce to a particular view of how selection influences the projection of category information from daughter nodes to their mothers. I show that by examining the nature of selection and projection, we can leverage the architecture of grammar to predict the requirement for wh-movement to be successive-cyclic: the projection rule makes it so that wh-phrases create their own barriers for extraction if their wh-features get too high, meaning they have to move outside the scope of their own features in order to extract. The theory entails that movement must be successive cyclic, but does not say through which positions. By varying the different allowed parameters in this theory, I show that it also captures variable sensitivity to the Keenan & Comrie hierarchy. Thus, the various locality requirements governing wh-movement can be reduced to basic principles governing selection and projection.

Speaker: Yizhen Jiang (MIT)
Title: Putting bare plurals into context (joint work with Yasu Sudo)
Time: Wednesday, October 25th, 1pm - 2pm
Location: 32-D461

Abstract: Bare plurals give rise to plurality inferences in positive sentences but not in negative sentences. There are two main approaches to this phenomenon. The implicature approach derives plurality inferences via scalar implicatures (Ivlieva 2020, Spector 2007, Sudo 2023, Zweig 2009, a.o.). The homogeneity approach attributes plural interpretations to trivalent semantics of bare plurals and claims a parallelism between bare plurals and plural definites (Križ 2017). Specifically, these two approaches make divergent predictions regarding the availability of plurality inferences in negative sentences and their sensitivity to context w.r.t. sentence polarity. We report on three experiments investigating plurality inferences of bare plurals via precise manipulation of context.

Our results show: 1) Plurality inferences are available in simple negative sentences (Exp 1) and quantified negative sentences under proper context (Exp 2); 2) the context sensitivity of plurality inferences exhibits a symmetry w.r.t sentence polarity in simple sentences (Exp 1) but an asymmetry in quantified sentences (Exp 2); 3) Partial plurality inferences are available in both positive and negative sentences but are insensitive to contextual manipulation. These results pose challenges to both the implicature and homogeneity approach. We discuss possible directions for both approaches to address these issues.

Speakers: Keely New, Premvanti Patel, Giovanni Roversi, Kate Kinnaird, Athulya Aravind (MIT)
Title: BUCLD Practice Talks
Time: Thursday, October 26th, 12:30pm - 2pm
Location: 32-D461

How toddlers answer multiple wh-questions (Keely New, Premvanti Patel and Athulya Aravind)

Abstract: The ability to comprehend wh-questions is one that already emerges in infancy. Infants comprehend subject whquestions like “Who ate something?” by 15-months, and distinguish them from object wh-questions like “What did Mom eat?” by 20-months. Adult linguistic competence, however, includes more complex wh-questions like “Who ate what?”, which in turn demand more complex answers. In this work, we investigate 2-and-3-year-olds’ understanding of multiple wh-questions. To do so, we probed their sensitivity to restrictions on which wh-word in a multiple question can be fronted (the “superiority constraint”). Using a novel “fly-in-the-wall” paradigm, which recreates naturalistic parent-child interactions, we elicited responses from toddlers on grammatical and ungrammatical multiple wh-questions, as well as single wh-questions. Toddlers differentiated single and multiple wh-questions, often giving adult-like pair-list responses to the latter. Both groups distinguished between ill-formed and well-formed questions in their response patterns, albeit in different ways.

Acquisition of *ABA paradigms in a child Artificial Language Learning Experiment (Giovanni Roversi, Kate Kinnaird and Athulya Aravind)

Abstract: Across the world’s languages, only certain types of linguistic patterns are attested, while other, equally logically possible patterns appear to be non-existent. Are these gaps accidental or do they in fact reflect biases in our linguistic system? We bring developmental data to bear on this issue. As a concrete case study, we examine irregular adjectival degree paradigms. For example, English has adjectives like good-better-best, and other types of irregular paradigms are known from other languages, but no language seems to have paradigms that would look like goodbetter-goodest. We designed an Artificial Language Learning study to determine whether unattested adjectival paradigms are in fact harder to learn than attested ones. Our preliminary results indicate that children indeed find the unattested paradigms harder to learn than the attested ones. This brings suggestive evidence to the idea that language learning might be constrained in such a way that excludes these paradigms.

Speaker: Giovanni Roversi (MIT)
Title: Ā-extraction, Word Order, and Object Shift (?) in Äiwoo
Time: Tuesday, October 17th, 1pm - 2pm
Location: 32-D461

Abstract: I will present some ongoing work in progress I’ve been doing on the clausal syntax of Äiwoo, an Austronesian language from the Solomon Islands. Largely, Äiwoo shows some reassuringly familiar Austronesian fare: the classic “pivot-only” restriction on Ā-extraction is respected, whereby only the agent can be extracted from Actor Voice clauses, only the theme from Undergoer Voice clauses, etc. Part of this talk will focus on a series of interesting exceptions to this restrictions. In particular, we’ll take a look at the complex correlation between different types of objects (DPs, null pronouns, overt pronouns, reflexive anaphors), clausal word order, and extraction possibilities. This will then lead us to reconsidering a known aspect of Austronesian voice, across many languages. A definite theme is normally banned in Actor Voice, and requires Undergoer Voice instead (“John built.{✗AV, ✓UV} this house”); however, this ban is lifted when Actor Voice is forced by external syntactic pressures, for example by Ā-extraction of the agent (“I met the man [who built.{✓AV, ✗UV} this house]”). Although this is a known fact in the Austronesian literature, and it’s easy enough to describe, I suggest it’s been underrated how complex it is to actually account for and implement it, and how a correct analysis of it might have bigger-picture implications for the syntax-semantics interface and the architecture of the grammar.

Speaker: Lucas Champollion (NYU)
Title: The Limits of Possible Worlds Semantics
Time: Friday, October 20th, 3:30pm - 5pm
Location: 32-141

Abstract: The standard paradigm of possible world semantics fails to account for the subject-matter preserving nature of entailment and for discrepancies between connectives in propositional logic and in natural language. These discrepancies show up in empirical domains such as imperatives and counterfactuals. This calls for a paradigm shift towards a hyperintensional framework. I adopt unilateral truthmaker semantics in the sense of Fine (2017), with a concomitant move from intersective to collective conjunction. I conceptualize truthmakers as Davidsonian events. This yields a natural conception of negative events. I sketch an application to perception reports. (Partly based on joint work with Timothée Bernard.)

Hadas Kotek will be participating in a panel on careers in AI/ML on campus. It will take place Thursday 10/19, 4-6pm, at 1-190. 

Speaker: Anastasia Tsilia (MIT)
Title: The future in desire: the case of Colloquial Jakartan Indonesian
Time: Wednesday, October 11th, 1pm - 2pm
Location: 32-D461

Abstract: Colloquial Jakartan Indonesian behaves like a tenseless language, with no tense morphology on the verb stem; the context disambiguates between a present and a past tense interpretation. As in many tenseless languages (Bochnak, 2019), the future is obligatorily marked by an adverb or akan/bakal will'. We argue that mauwant’ can be used to mark the future as well, as in the following:

(1) Context: We are at a party, but it’s getting late. I need to leave. Sebenarnya aku nggak mau, tapi aku mau pulang sekarang ya. Actually I neg want but I fut go-home now ok `I don’t actually want to but I will go home now.’

We call this use of mau the future mau. Future mau can have a purely temporal use, is compatible with inanimate subjects, and with the negation of mau meaning want' (also spelled out as pengen). However, future mau cannot be directly negated. It is incompatible with clausemate negation, negative quantifiers, and with the implicit negation triggered by the alternatives ofonly’ (Rooth, 1985). Yet, it is compatible with negation in a higher clause, as well as with negation in yes/no questions. It thus seems that we cannot negate future mau directly, but we can negate the proposition that contains it. We discuss the empirical picture, as well as argue that there is a dispositional requirement associated with future mau, namely a requirement that the subject is disposed to causing the eventuality of the verb. Finally, we propose that mau as a modal (Sneddon, 2010; Jeoung, 2020) is the dispositional will (Copley, 2002), having the meaning of bakal will' enriched with a presupposition about the dispositions of its subject. We also propose a solution to the negation puzzle, arguing that future mau behaves as a PPI, and following New (2023) in positing two adjunction sites for auxiliaries, one below and one above negP; being a PPI, future mau can only attach to the higher adjunction site. All in all, we show that mau in Indonesian can either quantify over buletic alternatives or simply over accessible worlds likewill’, with its desire' component being turned into a dispositional presupposition, eliminating the need for an attitude holder. Indonesian shows thatwant’ can synchronically mean `will’, a change which is diachronically attested in many languages (Heine, 2017).

Speaker: Jon Rawski (MIT)
Title: A Computational Puzzle from Signed Reduplication
Time: Thursday, October 12th, 12:30pm - 2pm
Location: 32-D461

Abstract: Reduplication is a cross-linguistically common yet computationally complex morphological copying process. Reduplication is far more ubiquitous and expressive in sign than in speech, regularly exhibiting partial and total copying, as well as triplication and other variants which are either very rare, unattested, or just impossible in spoken languages. Computationally, reduplication possesses linear growth (i.e. at most n copies, independent of the base form’s size), restricting it to the class of Regular functions in both weak and strong generative capacity (Rawski et al 2023, Dolatian et al 2021). This talk examines a unique signed phenomenon of “embedded” aspectual reduplication (Klima & Bellugi 1979, Wilbur 2009), where multiple copying processes compose within one another. This process is puzzling because it potentially allows polynomial growth, violating linearity. At issue is 1) whether regular functions are a necessary and sufficient condition for morphological computation, and 2) how representations and computations differ across speech and sign. I will argue that this embedded reduplication is still regular, since those functions are closed under composition. However, this depends on thorny issues like cyclicity, bounds on the number of copies, and how speakers vs signers represent the semantically-based input to reduplication. I will compare to spoken cases of repeated reduplication in Guébie, Tigre, and Runyankore, and discuss these issues.

Speaker: Tanya Bondarenko (Harvard)
Title: Scope-marking in Georgian, and the impossibility of long-distance movement
Time: Tuesday, October 3rd, 1pm - 2pm
Location: 32-D461

Abstract: In this work in progress, I present some observations about long-distance dependencies in Georgian. Georgian is a language that lacks long-distance wh-movement, (1): wh words cannot be extracted across a finite clause boundary. To express the meaning of a long-distance question, a construction with wh scope marking is used, (2), where a wh-expression in the embedded clause co-occurs with the wh-word ra “what” in the matrix clause.

(1) ​ Long-distance wh extraction
*ra-s1       pikrobs  mariami,       [rom     t1  č’ams      šota]?
what-ACC thinks  Mariam.NOM   COMP     is.eating    Shota.NOM
‘What does Mariam think that Shota is eating?’

(2) ​ Wh scope marking
ra-s       pikrobs  mariami,       [rom     ra-s        č’ams    šota]?
what-ACC thinks  Mariam.NOM   COMP  what-ACC  is.eating  Shota.NOM
‘What does Mariam think that Shota is eating?’

In this talk I argue that the construction in (2) does not involve a long-distance dependency: the two wh-expressions are not part of the same movement chain. I propose that a version of the indirect dependency approach (Dayal 1994) can account for the properties of wh scope marking in Georgian: the two wh-words undergo movement in each of their respective clauses to a position between VoiceP and TP (Borise 2023), and then the embedded CP is adjoined to the matrix one.

If this analysis is on the right track, it suggests that true long-distance dependencies are absent in Georgian, and the question is: why? At the end of the talk, I will offer some speculations I have, and solicit suggestions on where to look for the answer.

Speaker: Bingzi Yu (MIT)
Title: Studying naturalness bias in transmission and communication with ALL experiments
Time: Monday, October 2nd, 5pm - 6:30pm
Location: 32-D831

Abstract: Phonetic substance is believed to affect phonological grammar through acquisition. Patterns with a clear phonetic motivation are considered more natural and also typologically more frequent. While the hypothesis suggests a bias towards more natural patterns, experimental evidence has been inconsistent. In this talk, I will report the results of two Artificial Language Learning (ALL) experiments. With vowel harmony (natural) and vowel disharmony (unnatural) as the target patterns, the two experiments investigated naturalness bias in the context of transmission and communication. The first experiment found a similar decreasing tendency of both patterns over time, whereas the second one revealed a bias effect during interaction. The mixed results show the weakness of naturalness bias, in line with previous work, and highlight the impact of communicative context on the functioning of bias.

Speaker: Natalie Weber (Yale University)
Title: Resolving prosodic structure inside of polysynthetic words
Time: Friday, October 6th, 3:30pm - 5pm
Location: 32-141

Abstract:

Background
Over the last decade there has been a renewed interest in domain-delimited phonological processes and their correspondence with syntax (cf. Selkirk 2011; and overviews in Bennett & Elfner 2019; Elfner 2018; Elordieta 2008, Scheer 2011). Mismatches between phonological domains and syntactic constituents constitute a strong argument for an indirect theory of phonology, where phonological processes are delimited by a prosodic structure which is distinct from syntactic structure (Downing 1999; Hayes 1989; Inkelas 1993; Nespor & Vogel 2007/1986; Pierrehumbert & Beckman 1988; Selkirk 1986; Selkirk 1984). However, even within this framework there are open questions about the visibility and directionality of syntactic and prosodic constituents. This talk addresses two:

1. Modularity of syntax and phonology? Many recent theories allow morphosyntactic and prosodic representations to correspond within the same computational architecture (e.g. Alignment Theory: McCarthy & Prince 1994, Selkirk 1996, Werle 2009; Wrap Theory: Kabak & Revithiadou 2009, Truckenbrodt 1996, 1999; Match Theory: Elfner 2012, Selkirk 2011). Other theories require some amount of modularity (e.g., Scheer 2011; MSO-PI-PO: Itô & Mester 2023; Lee & Selkirk 2023).
2. Serial or parallel resolution of prosodic structure? Many recent theories implicitly or explicitly allow phonological derivation across a phonological representation which includes all prosodic categories at once (e.g., most theories of Prosodic Phonology, Nespor & Vogel 2011/1986, Selkirk 2011). Other theories require serial computation of prosodic categories from the innermost to the outermost (e.g., Cophonologies by Phrase: Sande et al. 2020; Stratal OT: Bermúdez-Otero 1999, Kiparsky 2000, 2008).

Prosodic structure within Blackfoot
This talk focuses on phonological processes within the verb in Blackfoot (Algonquian), a polysynthetic language. I argue that prosodic structure in Blackfoot shows that modularity and serial derivation must be maintained. The analytic framework I adopt is similar in some ways to Stratal OT (Bermúdez-Otero 1999, Kiparsky 2000, 2008), but without the same assumptions of how morphosyntactic structure maps to prosodic structure. The reason is that there is independent syntactic evidence in Blackfoot that the stem and the verbal complex are constructed in phrasal syntax. This in turn suggests that prosodic structure “within the word” may arise from the same kinds of correspondences with syntax that occur at the phonological phrase level.

A very warm welcome to ling-23 and visitors! 

ling-23

Soledad Chango: My name is Soledad Chango and just joined the MITILI program. I’m from an indigenous Kichwa community called Salasaka which is located in the highlands of Ecuador. I enjoy swimming, dancing, embroidering, and playing the violin a lot.

Cooper Roberts: I am Cooper Roberts, a PhD student interested in semantics and (morpho)syntax. Last spring, I graduated from Boston University across the river with a BA/MA in linguistics. Before moving to Boston, I lived in Pittsburgh, Pennsylvania. When I am not thinking about the semantics of classifiers or internal reduplication, I enjoy things like gardening, swimming, writing music, and watching hockey. I also split a one-bedroom apartment with three cats, so there’s never a dull moment. I look forward to getting to know everyone!

Yizhen Jiang: Hi everyone :) I’m Yizhen. I’m from Zhejiang, China. My main interests lie in semantics and pragmatics, combining formal and experimental methods. Outside linguistics, I enjoy Gomoku (too addictive that I uninstall but shortly reinstall the app frequently), K dramas (the sillier the better), and nature.

Hani Al Naaem: My name is Hani. I was born and raised in a small city to the north of Jordan. My interests within linguistics broadly lie in phonetics and phonology. More specifically, I’m interested in probing the atoms of phonological representations and in exploring linking hypotheses between those abstract representations and observed phonetic forms. I’m also interested in prosody/phrasal phonology and stress systems.
Outside linguistics, I enjoy watching sports (soccer and tennis mainly), going on walks, and just hanging out with friends. Spirituality/religion is also a very important part of my life; I’m a practicing muslim, and I’ve recently been dedicating more time to reading and studying the Qur’an more closely.

Paul Meisenbichler: My name is Paul Meisenbichler. I am from Vienna, and I completed my MA in linguistics at the University of Vienna. My main research interests are in semantics and syntax. My MA-thesis was concerned with the modal and temporal interpretation of determiner phrases. Outside of linguistics, I enjoy reading, watching movies/TV-series, and going to restaurants/bars with friends.

Dr. Jeffrey Toney (visiting professor): My faculty sponsor is Prof. Michel DeGraff and I’ve been contributing to the MIT Haiti Initiative. This past year, I have served as a New England Regional Consortium Fellow studying archives from eight institutions across the region, supported by the Massachusetts Historical Society. I am exploring ways to celebrate language, gender, culture and race, particularly to enhance STEM education. My UROP students were featured in an MIT News article, “Exploring culture, identity and the arts to enhance undergraduate education.” Last Spring, I retired from Kean University as Professor Emeritus and have been focusing on my research here at MIT, collaborating with colleagues across STEM, the arts and humanities.

Dr. Maya Honda (visiting professor): Hello, my name is Maya Honda. This is my second year as a visiting lecturer in the department. (I missed introducing myself last year.) My hometown is San Diego, California. My work focuses on linguistics education and pedagogy. I’ve worked around the world with teachers and with students of all ages to integrate the formal study of language and the achievements of linguistics into the school curriculum. I’m retired from the faculty of Wheelock College in Boston, where I taught linguistics, language development, and human development. I’ve come out of retirement to teach Linguistics in K-12 Education at MIT and to encourage everyone in the department to engage in linguistic outreach. (I’m that person who writes those emails!) Art museums are my go-to place here and wherever I happen to be.

Dr. Si Berrebi (visiting postdoc): I’m Si Berrebi, from Tel Aviv, mostly interested in questions about the representation, processing and social meaning of phonetic and phonological variation. I also appreciate coffee, walking, meditation and learning more about pedagogy.

Magdalena Lohninger (visiting student): Hi there, I’m Magdalena Lohninger (I usually go by the short version of my first name, Lena) and I’m a fourth year PhD student, visiting MIT during fall. Back home, I’m affiliated with the University of Vienna and work with Susi Wurmbrand. As a linguist, I’m interested in formal syntax and like to get lost on A’/A-movement paths. In my non-linguistic life, I try to cycle around the world (in small bits at a time) and climb onto things. I also like drinking beer, getting into unknown situations of whatever kind, sealions and poetry that doesn’t rhyme.

Dr. Jon Rawski (visiting professor): I am visiting from San Jose State University, where I am an Assistant Professor of Linguistics. I work in mathematical linguistics and learnability, placing mathematical boundary conditions on the grammars underlying human language, and on how they can be learned. I am currently investigating how the algebraic structure of a learner’s hypothesis space enables successful grammar acquisition, as an abductive inference problem. I also have a particular interest in signed languages, as they challenge theories of phonological and morphological cognition. This fall I am teaching a graduate seminar on Grammars and Representations, where we study classes of grammars and their consequences for linguistic theory.

Speaker: Tanya Bondarenko (Harvard)
Title: Scope-marking in Georgian, and the impossibility of long-distance movement
Time: Tuesday, September 26th, 1pm - 2pm
Location: 32-D461

Abstract: In this work in progress, I present some observations about long-distance dependencies in Georgian. Georgian is a language that lacks long-distance wh-movement, (1): wh words cannot be extracted across a finite clause boundary. To express the meaning of a long-distance question, a construction with wh scope marking is used, (2), where a wh-expression in the embedded clause co-occurs with the wh-word ra “what” in the matrix clause.

(1) ​ Long-distance wh extraction
*ra-s1       pikrobs  mariami,       [rom     t1  č’ams      šota]?
what-ACC thinks  Mariam.NOM   COMP     is.eating    Shota.NOM
‘What does Mariam think that Shota is eating?’

(2) ​ Wh scope marking
ra-s       pikrobs  mariami,       [rom     ra-s        č’ams    šota]?
what-ACC thinks  Mariam.NOM   COMP  what-ACC  is.eating  Shota.NOM
‘What does Mariam think that Shota is eating?’

In this talk I argue that the construction in (2) does not involve a long-distance dependency: the two wh-expressions are not part of the same movement chain. I propose that a version of the indirect dependency approach (Dayal 1994) can account for the properties of wh scope marking in Georgian: the two wh-words undergo movement in each of their respective clauses to a position between VoiceP and TP (Borise 2023), and then the embedded CP is adjoined to the matrix one.

If this analysis is on the right track, it suggests that true long-distance dependencies are absent in Georgian, and the question is: why? At the end of the talk, I will offer some speculations I have, and solicit suggestions on where to look for the answer.

Speaker: Hani Al Naeem (MIT)
Title: Timing and the Rhythm Class Hypothesis
Time: Monday, September 25th, 5pm - 6:30pm
Location: 32-D831

Abstract: The rhythm class typology classifies languages into stress-timed and syllable-timed, originally conceptualized in terms of the isochrony of stressed feet vs. isochrony of syllables, and later in terms of durational variability. Despite strong phonetic evidence against isochrony and inconsistent classification based on durational variability measures, the rhythm class typology enjoyed a lasting popularity in the literature. Arguably, the reason for this is the strong evidence for rhythm classes in perception research in the late 1990s and early 2000s. Recent research, however, provided counterarguments yet again. In this talk, I will (i) give a quick overview of the rhythm class hypothesis, (ii) report on a small project that I worked on using one type of durational measures, and (iii) discuss results from the literature that cast doubt on the usefulness of such measures and the rhythm class hypothesis itself.

Congratulations to Giovanni Roversi, whose joint paper with Sebastian Sauppe, Åshild Næss, Martin Meyer, Ina Bornkessel-Schlesewsky, and Balthasar Bickel was just published in Cognitive Science! Take a look at the paper, An Agent-First Preference in a Patient-First Language During Sentence Comprehension, here. 

Speaker: Keely New (MIT)
Title: Voice and the variable position of auxiliaries in colloquial Jakartan Indonesian
Time: Tuesday, September 19th, 1pm - 2pm
Location: 32-D461

Abstract: Colloquial Jakartan Indonesian (CJI) displays a voice system typical of Austronesian languages: in each clause there is one argument (the “pivot”) that occupies the sentence-initial position, and the choice of this argument affects verbal morphology as well as the word order of other arguments. The pivot is also the only argument that may undergo A’-extraction (the pivot-only restriction). In this talk, I discuss a potential violation of the pivot-only restriction in CJI and show how it presents a challenge for standard, locality-based approaches. I argue that this violation is only apparent by proposing that there are multiple position positions for auxiliaries in the structure.

Speaker: Hadas Kotek (MIT)
Title: Gender bias and stereotypes in Large Language Models
Time: Thursday, September 21st, 12:30pm - 2pm
Location: 32-D461

Abstract: In this talk I’ll discuss my recent co-authored paper on gender bias in Large Language Models (LLMs). We used syntactically ambiguous sentences that contained one stereotypically male occupation-denoting noun and one stereotypically female occupation-denoting noun along with a gendered pronoun in a simple 2x2 paradigm, directing the model to engage in a simple pronoun disambiguation task:

(1) a. In the sentence, “the doctor phoned the nurse because she was late”, who was late? b. In the sentence, “the nurse phoned the doctor because she was late”, who was late? c. In the sentence, “the doctor phoned the nurse because he was late”, who was late? d. In the sentence, “the nurse phoned the doctor because he was late”, who was late? (2) Could {“he”, “she”} refer to the other person instead?

We tested four recently published LLMs and observed gender bias such that LLMs are 3-6 times more likely to choose an occupation that stereotypically aligns with a person’s gender. These choices align with people’s perceptions better than with the ground truth as reflected in official job statistics. LLMs in fact amplify the bias beyond what is reflected in perceptions or the ground truth.

Importantly for us as linguists, the LLMs ignore crucial ambiguities in sentence structure 95% of the time in our study items, but when explicitly prompted as in (2), they are able to recognize the ambiguity. The LLMs further provided explanations for their choices that were factually inaccurate and likely obscure the true reason behind their predictions. Those explanations often relied on mischaracterizations of syntactic or semantic facts about the sentences or about linguistic theory, which I characterize as grammatical hallucinations. The talk will discuss ways in which these findings provide an opening for linguists to better educate the public and work to improve the data used to train LLMs.

A joint paper by David Pesetsky and Kanoe Evile, wh-which Relatives and the Existence of Pied-Piping has just appeared in Glossa. It originated as an Intro to Syntax squib by Kanoe, a Linguistics minor who graduated last spring, who is starting medical school. Congratulations Kanoe and David! 

Congratulations to Kai von Fintel and Sabine Iatridou! Their article, Publication of Prolegomena to a theory of X-marking, was published in Linguistics and Philosophy. It also received coverage on MIT News! 

Speaker: Giovanni Roversi (MIT)
Title: Binding and anti-cataphora in Äiwoo: domain-based restrictions on interpretation?
Time: Tuesday, September 12th, 1pm - 2pm
Location: 32-D461

Abstract: I report some work in progress, based on ongoing fieldwork, on how Condition C works in Äiwoo, an Oceanic (< Austronesian) language from the Solomon Islands. The central puzzle is that Condition C (or more generally, the mechanism that determines whether two nominals can or cannot be coreferent) shows a non-uniform profile in this language. In certain types of clauses one seems to get perfectly traditional Condition C effects, where the important factor is c-command between two nominals and not linear precedence (e.g. “His_i mother loves John_i”, where “his” precedes “John” but doesn’t c-command it, and coreference is possible). In other types of clauses, however, we see a completely different profile, where the (im)possibility of coreference seems to fully disregard structural factors (c-command) and only care about linear precedence. I will suggest a hypothesis whereby there’s a particular high domain in the Äiwoo clause which blocks cataphora, but not otherwise. This is very much work in progress, so I welcome any brainstorming and feedback!

Speaker: Johanna Alstott (MIT)
Title: Before and after decomposing first and last
Time: Thursday, September 14th, 12:30pm - 2pm
Location: 32-D461

Abstract: First and last have been variously described as ordinals (Bhatt 2006; Bylinina et al. 2014), superlatives (Sharvit 2010), or “both ordinals and superlatives” (Charnavel 2022). These descriptions are generally loose and undefended, and those who label first and last as superlatives do not present and argue for a particular decomposition. Thus, first and last’s status as ordinals vs. superlatives and their internal composition remain open issues. In this work, I argue that first and last are superlatives, in particular the superlative forms of before and after. As evidence that first and last are superlatives, I show that they pattern like superlatives and unlike ordinals (second, third, etc.) with respect to plurality, modifier choice, modal superlatives with possible, and the ordinal superlative construction. I next argue that the relations between before and first and between after and last show themselves overtly in paraphrases and the etymology of first; furthermore, first and last semantically differ in ways that before and after have also been noted to differ. Formalizing the proposed decomposition of first and last necessitates either a (non-standard) treatment of before and after as comparatives or a treatment of superlatives that is non-standard in semantics but standard in morphosyntax (Bobaljik 2012). I survey evidence that could adjudicate between the two strategies for decomposing first and last, arguing that the latter is more plausible.

Speaker: Tanya Bondarenko (Harvard)
Title: Scope-marking in Georgian, and the impossibility of long-distance movement
Time: Tuesday, September 5th, 1pm - 2pm
Location: 32-D461

Abstract: In this work in progress, I present some observations about long-distance dependencies in Georgian. Georgian is a language that lacks long-distance wh-movement, (1): wh words cannot be extracted across a finite clause boundary. To express the meaning of a long-distance question, a construction with wh scope marking is used, (2), where a wh-expression in the embedded clause co-occurs with the wh-word ra “what” in the matrix clause.

(1) ​ Long-distance wh extraction *ra-s1       pikrobs  mariami,       [rom     t1  č’ams      šota]? what-ACC thinks  Mariam.NOM   COMP     is.eating    Shota.NOM ‘What does Mariam think that Shota is eating?’

(2) ​ Wh scope marking ra-s       pikrobs  mariami,       [rom     ra-s        č’ams    šota]? what-ACC thinks  Mariam.NOM   COMP  what-ACC  is.eating  Shota.NOM ‘What does Mariam think that Shota is eating?’

In this talk I argue that the construction in (2) does not involve a long-distance dependency: the two wh-expressions are not part of the same movement chain. I propose that a version of the indirect dependency approach (Dayal 1994) can account for the properties of wh scope marking in Georgian: the two wh-words undergo movement in each of their respective clauses to a position between VoiceP and TP (Borise 2023), and then the embedded CP is adjoined to the matrix one.

If this analysis is on the right track, it suggests that true long-distance dependencies are absent in Georgian, and the question is: why? At the end of the talk, I will offer some speculations I have, and solicit suggestions on where to look for the answer.

The 33rd meeting of Semantics and Linguistic Theory (SALT 33) was hosted by the linguistics department at Yale University on May 12-14, 2023. MIT was represented by several current students and alumni:

• Johanna Alstott (first-year): Ordinal Numbers: Not Superlatives, but Modifiers of Superlatives
• Ido Benbaji (fourth-year): Is nothing irreplaceable? A substitution theory of de re
• Omri Doron (fourth-year): Another look at the Mapping Hypothesis: evidence from Hebrew 
• Yash Sinha (fourth-year): Number morphology in Hindi coordinative compounds
• Tuệ Trịnh (PhD, 2011): Contextual bias anti-licenses NPIs in polar questions
• Deniz Özyıldız & Wataru Uegaki (PhD, 2015): Two kinds of question-embedding strategies & veridicality alternations
• Luis Alonso-Ovalle & Aaron Hirsh (PhD, 2017): Weakening is external to ‘only’

Speaker: Ksenia Ershova (MIT)
Title: Move me high, pronounce me low: Diagnosing argument asymmetries in West Circassian nominalizations
Time: Tuesday, May 16th, 1pm - 2pm
Location: 32-D461

Abstract: Nominalizations in West Circassian, a polysynthetic, high absolutive language, display a puzzling mismatch in diagnostics for argument asymmetries. On the one hand, reciprocal binding suggests that the theme of a transitive verb c-commands the external argument, analogous to finite clauses. On the other hand, the ordering of arguments in nominalizations are constrained in accordance with thematic roles, with the external argument obligatorily surfacing further from the verbal root than the internal argument, suggesting that the external argument c-commands the theme. I argue that this apparent mismatch results from constraints on licensing and spell-out in the nominal domain: while the internal argument raises to a position c-commanding the ergative agent, it must be spelled out in the lower position due to conditions on licensing by adjacency. The analysis resolves an otherwise puzzling contradiction between structural diagnostics and confirms that syntactic licensing constraints may inform PF spell-out in ways which obscure the underlying syntactic structure.

Speaker: Mitya Privoznov (MIT)
Title: Coordinate Structure Constraint, CED, and Spell Out
Time: Tuesday, May 9th, 1pm - 2pm
Location: 32-D461

Abstract: It has been argued that in a coordinate construction, as in (1), the connective forms a constituent with the second conjunct: [X [and Y]] (see, e.g., Ross, 1967; Pesetsky, 1982).

(1) I [went home [and ate dinner]].

It has also been argued that the connective is the head of the coordinate structure (see, e.g., Melčuk, 1974; Paducheva, 1974; Paducheva and Zaliznyak, 1979; Munn, 1987, 1992, 1993; Larson, 1990; Johannessen, 1993, 1998; Citko, 2005; Zhang, 2010). In other words, in (1), when and is merged with ate dinner, and projects; and when went home is merged with and ate dinner, the latter projects. Thus, we arrive at the ConjP/CoP/&P structure with the second conjunct being the complement of and (Conj, Co, &) and the first one being a specifier of ConjP/CoP/&P. At the same time, since Ross (1967), it has been known that in most cases, in a coordinate structure, nothing can be extracted from either conjunct (Coordinate Structure Constraint, CSC). At the same time, Ross (1967) already points out certain systematic exceptions to CSC, which have been further discussed by Goldsmith (1985), Lakoff (1986) and Postal (1998), among others.

In this talk, I will propose a preliminary version of an analysis of coordinate structures like (1) which involves structural ambiguity. More precisely, I suggest that coordinate constructions, as in (1), are, in fact, ambiguous between three parses: (a) [XP=YP XP YP] structure (where both XP and YP are maximal projections); (b) [XP X’ YP] structure, originally proposed by Postal (1998), where the second conjunct is an adjunct to the first; and (c) ConjP/CoP/&P structure. Notably, the syntactic category (and the lexical semantics) of and is the same across all the three parses. The ambiguity is syntactic, though it may have pragmatic effects (cf. Bjorkman 2010, 2013, 2014, Bassi and Bondarenko 2021). I will present some preliminary evidence from Russian, which shows certain advantages of such an account in explaining: (a) the overt morphology of coordinate structures in Russian and potentially cross-linguistically; (b) the extraction properties of coordination and the “violable” status of CSC; (c) the presupposition projection profile of conjunction; (d) the distribution of overt connectives and group readings in Russian DP-conjunctions.

Speaker: Suzana Fong (Yale)
Title: A Wholesale Late Merge Theory of Control
Time: Thursday, May 11th, 12:30pm - 2pm
Location: 32-D461

Abstract: Wholesale Late Merge (Takahashi & Hulsey 2009) consists in countercyclically merging the restrictor with an operator that had been previously merged into the derivation. The Movement Theory of Control (Hornstein 1999), in turn, is a theory of control according to which there is no dedicated control module in the grammar; rather, obligatory control is the byproduct of moving a DP through more than one thematic position. In this nascent project, I entertain the consequence of combining these two independent theories. The result can be dubbed ‘Wholesale Late Merge Theory of Control’, whereby obligatory control is the consequence of early merging a D in the embedded clause, moving it into a thematic position in the matrix clause, upon which point, an NP is countercyclically WLM-ed. If correct, this theory accounts for three asymmetries between raising and control that are not expected in the simplest version of the MTC, namely, (i) the fact that control, unlike raising, does not reconstruct, (ii) the lack of case connectivity in Icelandic control, and (iii) the enforcement of anti-pronominality effects in control, which are absent in raising. Additionally, the early merged D provides an apt model for instances of obligatory control PRO that are realized as an overt pronoun.

Students in 24.S95 Linguistics in K-12 Education will make public presentations about their work on Wednesday, May 10^th, 10:30 am – 12:00 pm in the 8^th Floor Conference Room 32-D831. To accommodate people’s schedules, they will repeat their presentations later the same day from 3:00 pm – 4:30 pm, in 26-142. 

The students will talk about their diverse experiences designing and teaching linguistics lessons for middle school and high school students. 

The 41st West Coast Conference on Formal Linguistics (WCCFL 41) was held at the University of California, Santa Cruz on May 5-7, 2023. The following members of our community presented at the conference:

• Haoming Li (1st year), Zhouyi Sun (1st year): Mandarin Clausal Comparatives Involve Standard Embedding
• Giovanni Roversi (3rd year): Adjectival “concord” in North Sámi is not concord (and it’s two different phenomena)
• Janek Guerrini (visiting student): Kind predication, flavors of genericity, and cumulativity

A number of alums-friends also gave talks and posters:

• Michela Ippolito (PhD, 2002), Angelika Kiss, Will Williams: Discourse only
• Michela Ippolito (PhD, 2002), Zahra Mirrazi: Modal Past is Past: Evidence from non-SOT langauges
• Seth Cable (PhD, 2007), James Crippen: Stative marking in Tlingit: Evidence from the complexity of states
• Andrew Hedding, Michelle Yuan (PhD, 2018): Phase unlocking and the derivation of verb-initiality in San Martin Peras Mixtex
• Ken Hiraiwa (PhD, 2005), Kimiko Nakanishi: Disjunction as question: Disjunction is not a PPI in Japanese
• Ivona Kučerová (PhD 2007) and Alan Munn: Beyond ϕ-features: Are we there yet? Agree reconsidered
• Colin Davis (PhD, 2020): Anti-locality explains the restricted interaction of subjects and parasitic gaps
• Colin Davis (PhD, 2020), David Diem: Doubling by movement within and from PP in Alemannic German 
• Yağmur Sağ, Ömer Demirok (PhD 2019), Muhammet Bal: Subject pseudo-incorporation in Laz
• Abdul-Razak Sulemana (PhD, 2021): Passive without morphology: a case for implicit arguments