MIT Linguistics

09/29/2026

Colloquium - Philippe Schlenker (Institut Jean-Nicod, CNRS)

Speaker: Philippe Schlenker (Institut Jean-Nicod, CNRS)
Title: "The Iconic Frontier of Formal Grammar"
When: Friday, October 2nd, 3:30-5pm
Where: 32-141

Abstract:
Iconicity has long been known to be pervasive in sign language. But how does it interact with grammar? On the basis of ASL data, we propose a modular framework with three key features.

- Semantics: Logical semantics must be extended with an iconic component. The result, which we call “Iconological Semantics”, has Logical Forms that combine standard logical operators with an iconic component (inspired by formal pictorial semantics). The “glue” between them is viewpoint manipulation.

- Syntax: Certain expressions with a designated iconic behavior, called ‘classifiers’, have not just an iconic semantics but also an iconic syntax: they take their arguments in the order in which they would appear in a simplified comic. To say that “a crocodile ate up a ball”, the normal word order in ASL is SVO if the verb (EAT) is not a classifier. But if one uses a classifier instead, the order becomes SOV or OSV. This is because one typically sees the eating after one has seen the event participants.

- Pragmatics: The iconic content of classifiers is not an unstructured whole. It is divided among traditional slots of the inferential typology, such as assertions, presuppositions, implicatures, and more. Since this iconic content is determined by classifier movement and could not be listed in the lexicon, there must be an algorithm that productively divides this iconic content among these slots.(These claims extend from sign with iconicity to speech with gestures, but for reasons of time we will focus on sign.)

Brief survey: "Meaning after the Multimodal Revolution: Seven Things Signs and Gestures Teach Us about Meaning", to appear in WIREs Cognitive Science, https://ling.auf.net/lingbuzz/010015/current.pdf?_s=IFTGTQwFgPVfgVWK

http://whamit.mit.edu/2026/09/28/colloquium-phili…-jean-nicod-cnrs/

09/29/2026

LFRG - Prarthanaa Bharadwaj (Goethe University, Frankfurt)

Speaker: Prarthanaa Bharadwaj (Goethe University, Frankfurt)
Title: Polyfunctionality of the Particles -oo and -uu: The Indeterminate sytem in Kannada
Time: Wednesday, September 30, 1-2pm
Location: 32-D461

Abstract: In his seminal typology, Haspelmath (1997) observed that indefinite pronouns in many languages are built from interrogatives combined with focus-sensitive particles, and that the same building blocks yield negative polarity items (NPIs) and free choice items (FCIs). Work on such indeterminate systems has shown that they also extend to question formation (Kuroda, 1965; Kratzer & Shimoyama, 2002; Cable, 2010) and relative clauses (Slade, 2011; Dayal, 1997; Caponigro, 2003; Balusu, 2021). In this talk, I examine indeterminates in Kannada (Dravidian) from the perspective of the particles that license them. I present the paradigm of the disjunctive particle -oo and the inclusive particle -uu across coordination, indefinites, questions, inclusive focus, and relative clauses, and show that both the presence of a wh-word and the particle’s height of attachment determine its interpretation: -oo can occupy the clause-typing position, where it forms questions, whereas -uu never can. I then assess how far two existing accounts—Uegaki’s (2018) unified semantics of the Japanese Q-particle ka and Balusu’s (2022) analysis of Dravidian NPIs/FCIs as unconditionals can jointly account for the Kannada paradigm.

http://whamit.mit.edu/2026/09/28/lfrg-prarthanaa-…ersity-frankfurt/

09/29/2026

Syntax Square 10/29 - Magdalena Lohninger (Salzburg)

Speaker: Magdalena Lohninger (University of Salzburg)
Title: Form of a CP, meaning of a vP: Event complements in Austrian German
Time: Tuesday, September 29, 1:00pm - 2pm
Location: 32-D461

The talk investigates clausal complementation in Austrian German, focusing specifically on complement clauses under verbs like “try, begin, finish”. While such verbs select only infinitival complements in Standard German, they can take full CPs in Austrian German, giving rise to sentences such as (1):1) (1) Er hat versucht/angfangen/aufghört [ dass er des Auto repariert ] He has tried/started/finished [ that he the car repair.3sg ] He tried/started/finished to repair the carEven though such complement clauses surface as full CPs, they are highly restricted in their interpretation: they exhibit obligatory control, their time-reference must be identical to the matrix time, they cannot express any other morphological tense value than present tense, they can receive an implicative reading under verbs like “forget”, and do not permit embedded V2 or interrogatives. Hence, their interpretation resembles an infinitive (or what Wurmbrand and Lohninger 2019 term an “Event” clause), rather than what we expect from a CP. Interestingly, these “fake” CPs freely alternate with genuine infinitival complements, raising a puzzle concerning the relationship between semantic selection and syntactic size.Given that this is very recent work, I will focus primarily on the empirical findings, and hope to get some feedback on how to develop the more theoretical aspects.

http://whamit.mit.edu/2026/09/28/syntax-square-ma…hninger-salzburg/

The scripts.mit.edu service is currently unavailable. The service maintainers are working on the problem and expect to have it resolved soon.

09/21/2026

LingLunch - Eyal Marco (MIT), Roni Katzir (Tel-Aviv University), Ezer Rasin (Tel-Aviv University)

Speaker: Eyal Marco (MIT), Roni Katzir (Tel-Aviv University), Ezer Rasin (Tel-Aviv University)
Title: The complementary-distribution challenge for phonological learning: The case of [h] and [ŋ] in English
Time: Thursday, September 24, 12:30pm - 2pm
Location: 32-D461

Abstract: The distribution of [h] and [ŋ] in English poses a well-known challenge for the structuralist assumption that segments in complementary distribution are allophones of the same phoneme. In this study, we show that this distribution also poses a challenge within generative linguistics, a challenge of explanatory adequacy. The challenge is to explain why the English-learning child does not acquire a grammar in which [h] and [ŋ] are derived from the same underlying representation. The generative literature has not identified this as a problem, largely because of the assumption that [h] and [ŋ] are too phonetically dissimilar to participate in an alternation. However, we argue that phonetic accounts do not eliminate the challenge. We discuss simplicity-based theories of learning and show that under naive representational assumptions, simplicity-based learners prefer to derive [ŋ] from an underlying /h/, rather than a longer sequence of /ng/ (or an underlying /ŋ/), and therefore make wrong predictions about English speakers’ knowledge. The same learners, however, make correct predictions when combined with a representational framework that assumes underspecification and lexical redundancy rules. We therefore conclude that the acquisition problem can be addressed using simplicity-based learning, and that the complementary distribution of [h] and [ŋ] provides an argument for underspecification and lexical redundancy rules.

Speaker: Eyal Marco (MIT), Roni Katzir (Tel-Aviv University), Ezer Rasin (Tel-Aviv University)Title: The complementary-distribution challenge for phonological learning: The case of [h] and [ŋ] in EnglishTime: , 12:30pm - 2pmLocation: 32-D461Abstract: The distribution of [h] and [ŋ] in English pose...

09/21/2026

LFRG - Zachary Feldcamp (MIT)

Speaker: Zachary Feldcamp (MIT)
Title: Generalization X under Movement
Time: Wednesday, September 23, 1-2 p.m.
Location: 32-D461

Abstract: Noun phrases may often be interpreted either de re or de dicto with respect to an intensional operator, while predicates are obligatorily interpreted de dicto (Percus’s 2000 Generalization X). Existing accounts capture this asymmetry by constraining the relation between a predicate’s LF position and its intensional interpretation. I investigate what happens when a predicate overtly raises across an intensional operator in “predicate inversion” constructions such as With Taylor is believed to be a lawyer. I show that such predicates remain obligatorily de dicto, even though independent binding and Condition C diagnostics indicate that they need not syntactically reconstruct. This creates a reconstruction paradox: the raised predicate must be interpreted upstairs for some purposes but downstairs for intensionality. The result poses a problem both for scope theories and for theories that posit a predicate-internal intensional variable. I argue that the data instead favor a treatment of intensionality in which predicates take an intensional argument, and that, if this argument is saturated by a variable, such a variable must compose above the predicate constituent. I then explore how such an analysis can be implemented through semantic reconstruction while preserving the familiar de re/de dicto flexibility of nominal arguments.

Speaker: Zachary Feldcamp (MIT) Title: Generalization X under Movement Time: Wednesday, September 23, 1-2 p.m. Location: 32-D461 Zoom link: https://mit.zoom.us/j/92053685079 Abstract: Noun phrases may often be interpreted either de re or de dicto with respect to an intensional operator, while predic...

09/21/2026

Phonology Circle - Christopher Legerme (MIT)

Speaker: Christopher Legerme (MIT)
Title: Representation and Computation in Phonological Theory: Faust and Scheer (2025)
Time: Monday, September 21, 5pm - 6:30pm
Location: 32-D831

Abstract: Before asking what rules or constraints a grammar needs, we need to know what kind of representations those computations should be operating over in the first place: “If the representations are right, then the rules will follow” (McCarthy 1988:84). I’ll be presenting Faust & Scheer’s (2025) work on the behavior of stem-final TR clusters in North-Eastern Creusois (NEC), using the paper to foreground a broader methodological question pertaining to the division of labor between representation and computation in phonological theory. NEC shows a noun/verb asymmetry in the repair of illicit stem-final TR clusters. The account of Faust & Scheer combines Strict CV representations with Optimality Theory (OT) to derive deletion in nouns and metathesis in verbs through non-transderivational morphological exponence that is followed by a completely regular phonological computation in which phonology doesn’t need to make reference to syntactic category (2025: 32). And so, we are brought to consider the broader architectural question of what a modular grammar might require of its interfaces. For Faust & Scheer, their study of the NEC pattern is only one contribution to a larger debate over whether phenomena that appear to require cross-modular information really force us to weaken certain assumptions about modularity, or whether they can instead be derived by reconsidering our phonological representations and how the different components of grammar get related to one another. Please note that, as the inaugural talk of the 2026-27 season of Phonology Circle, this presentation will be oriented primarily toward introducing the NEC data, which instantiate the kinds of patterns familiar from morphophonology more generally, and the basic concepts needed to follow the paper’s argument, including the relevant machinery of OT, Strict CV, and some background assumptions about Spell-Out that motivate the analysis, rather than toward adjudicating the broader theoretical debate. That said, any comments, thoughts, or questions will of course be very welcome.

Speaker: Christopher Legerme (MIT)Title: Representation and Computation in Phonological Theory: Faust and Scheer (2025)Time: , 5pm - 6:30pmLocation: 32-D831Abstract: Before asking what rules or constraints a grammar needs, we need to know what kind of representations those computations should be ope...

09/21/2026

Saengthong accepted for publication by STAR!

A paper co-authored by our 3rd year student Rotsuprit (Ford) Saengthong and John Gluckman (University of Kansas), entitled “Thai Tough-Constructions and the Tough Signature”, has been accepted for publication by Syntactic Theory & Research (STAR). Below is the abstract:

We investigate the tough-construction in Thai, focusing on two exceptional properties. First, Thai tough-constructions may involve a potentially unconstrained dependency, and second, the gap in the embedded clause can be in the subject position. While these facts superficially support a heterogeneous view of tough-constructions across the world’s languages, on closer inspection, we come to a different conclusion. We argue that unconstrained dependencies and subject gaps are not a function of Thai tough-constructions, but instead reflect the fact that Thai has an undiagnosed broad subject position. Once controlled for, Thai tough-constructions look surprisingly like other well-documented cases. Ultimately, we are concerned not with variation in tough-constructions, but rather crosslinguistic similarities.

You can read the paper here: https://lingbuzz.net/lingbuzz/010319

Congratulations, Ford!

09/14/2026

Colloquium - Jonah Katz (UCLA)

Speaker: Jonah Katz (UCLA)
Title: "Template structure in musical textsetting: Evidence from early hip-hop"
When: Friday, September 18th, 3:30-5pm
Where: 32-141

Abstract:This talk describes the setting of lyrics to musical rhythms ('flow') in mid-1980s hip-hop. I argue based on the distribution of musical and linguistic features that the genre makes use of abstract templates, but that these templates are fundamentally different from metrical verse. Halle & Keyser (1971) propose that metrical poetry involves a mostly-invariant, abstract template of strong and weak metrical positions grouped into constituents of various sizes, and that metrical variation within a form stems from correspondence rules governing which linguistic materials may appear in which positions in the template. There has been significant disagreement about whether musical textsetting also involves such templates, and whether and how they might differ from poetic ones. For instance, Hayes and various colleagues (e.g. Hayes & MacEachern 1996, Hayes 2009) develop a model of folk-quatrain chanting and songforms where linguistic representations interact directly with musical meter and no abstract template mediates the relationship. Kiparsky (2006) argues that this quatrain form requires abstract poetic templates of the Halle & Keyser (1971) type to mediate between linguistic representations and musical meter. And Dell (2015) proposes, based on French song, that while textsetting does rely on templates, those templates are not abstract poetic ones: they instead encode information specific to musical tonality, rhythm, and meter.The current study offers evidence from hip-hop that tends to support Dell’s view broadly, but also requires significant extensions to his theory of representations. I show that this genre uses templates to enforce local parallelism, that the templates encode multiple hierarchical levels of prominence and constituency, and that the mapping of these rhythmic templates to musical metrical positions is partly arbitrary in ways that can’t be captured by global constraints on linguistic-metrical correspondence. All of these properties are analogous or identical to general features of music, including music without words (Lerdahl & Jackendoff 1983). I therefore propose that musical textsetting be modelled with Lerdahl & Jackendoff’s hierarchical template formalism (‘time-span reduction’) for music.

09/14/2026

Minicourse - Jonah Katz (UCLA)

Speaker: Jonah Katz (UCLA)
Title: "Synchronic lenition"
When: Wednesday, September 16th, 1pm-2:30pm (Day 1) + Thursday, September 17th, 12:30-2pm (Day 2)
Where: 32-D461

Abstract: In this mini-course, I'll give an overview of several different frameworks for thinking about synchronic lenition (and fortition) patterns, then review some of my own work in the area as it relates to these ideas. The first day will focus on what lenition is, and how linguists have characterized lenition patterns in terms of articulation, perception, and information. The second day will focus more narrowly on intervocalic lenition and initial strengthening, offering some case studies and (hopefully) new data from my lab.

09/14/2026

Syntax Square 10/15 - Giovani Roversi (MIT)

Speaker: Giovani Roversi (MIT)
Title: Applicative left behind?
Time: Tuesday, September 15, 1:00pm - 2pm
Location: 32-D461

I will present you some curious data from Äiwoo that I’ve been banging my head on for a while. We will see a construction that functionally speaking seems to be a relative clear case of an applicative (i.e., it adds one DP argument to the argument structure of the verb, one that would’ve otherwise had to be expressed as a PP). However, the way it’s morphologically realized is rather puzzling, because it’s nowhere near where applicatives typically are found. Given what they do syntactically, applicative heads are usually in the lower theta domain of the clause (i.e., within vP/VoiceP). Therefore, by the Mirror Principle (Baker 1988)’s terms and conditions, applicative morphemes should be very close to the verb stem — and indeed they are, in most (all?) cases I know of. However, in the Äiwoo case, the particle that marks this “applicative” construction is realized somewhere else entirely. I will discuss a couple of hypotheses about what we could be looking at, and where they break. Put on your creative syntactician hats, and help me out!

http://whamit.mit.edu/2026/09/14/syntax-square-10…vani-roversi-mit/