Whamit!

Speaker: Haoming Li & Zhouyi Sun (MIT)
Title: Mandarin clausal comparatives involve standard embedding
Time: Tuesday, April 25th, 1pm - 2pm
Location: 32-D461

Abstract: Erlewine (2018) argues that the Mandarin bi comparative is a form of clausal comparative. Since clausal comparatives are generally analyzed as involving degree abstraction (Bresnan 1973; Heim 2000; Bhatt and Pancheva 2004, a.o.), yet Mandarin is held by many to lack degree abstraction (Beck et al. 2009; Krasikova 2008; Erlewine 2018), a contradiction seems to arise. Erlewine reconciles the conflicting points by arguing that clausal comparatives are possible as long as we adopt a coordinate relation between the two clauses and employ ‘Degree Last’ gradable predicates, which he claims is precisely the case for Mandarin bi comparatives.

In this talk, we present several novel arguments in favor of the proposal that the syntactic structure of bi comparatives involve an embedding relation between the two clauses rather than a coordinate relation. We further illustrate the necessity of degree abstraction for an embedding approach. This lends support to the argument that Mandarin does indeed have degree abstraction (Gong and Coppock 2021), and casts doubt on the proposal of the ‘Degree Abstraction Parameter’ (DAP) (Beck et al. 2004).

Speaker: Anastasia Tsilia (MIT)
Title: Hidden causality in Modern Greek
Time: Wednesday, April 26th, 1pm - 2pm
Location: 32-D461

Abstract: We investigate certain Modern Greek attitudinal constructions, where an attitude verb takes an ACC DP as an argument, which is then referred to by a pronoun in the subject or the object position of the embedded clause. Here is an example of this construction:

(1) I Maria theli ton Yani_i [na pro_i aghapai mono aftin].
The.nom Maria.nom want.prs the.acc Yani.acc [subj pro love only her.acc] ‘Maria wants Yanis to only love her’

We argue that the ACC DP does not move from the lower clause (see Hadjivassiliou et al. (2000); Kotzoglou and Papangeli (2007); Kotzoglou (2013, 2017)), but is rather base-generated in a different clause. Given that it also has to co-refer with a pronoun in the subject or object position of embedded clause, we argue that this is an instance of prolepsis. Surprisingly, despite being seemingly outside of the scope of the attitude verb, the ACC DP can be read de dicto:

(2) Context: Little Petros is in kindergarten and he and his friends believe that green dogs exist. One day they are talking about green dogs and Petros bets that exactly three of them will show up at his party.

O Petrakis theli tris prasinus skilus [na erthun sto parti]. The.nom Petros.dim want.prs three green.acc dog.acc [subj come to-the party].  `Little Petros wants three green dogs to come to the party.’

On top of that, there is a semantic requirement that the ACC DP is part of the cause of the complement clause. We discuss this requirement and provide a clausal analysis, inspired by intensional transitive verbs (den Dikken et al.2018), arguing that there is a covert clause boundary between the matrix and the complement clause, featuring a weak causative CAUSE:

(3) …want [proleptic-DP CAUSE [CP…]]

Under such an analysis, the de dicto readings of the proleptic DP follow naturally, since it is in the scope of the attitude verb.

Speaker: Maria Giavazzi (DEC, ENS)
Title: Stress-conditioning from diachrony to synchrony: The case study of velar palatalization
Time: Monday, April 24th, 5pm - 6:30pm
Location: 32-D831

Abstract: Prosodically prominent positions within the word are privileged, manifesting the positional maintenance of contrasts otherwise neutralized and the resistance to processes applying elsewhere (a.o. Beckman, 1998; Giavazzi, 2010, González 2013). These positions are also the preferred target for a small class of frequent processes, e.g., consonants are often lengthened in pre-tonic and post-tonic position (Lavoie 2001, Smith 2002). In this talk I will focus on a case study of such a stress-conditioned process, velar palatalization in Italian (Giavazzi, 2010). First, I will present results of a study investigating the distribution of this process in early Italian. The application of palatalization in Early Italian was much more variable than in the contemporary language, though stress-conditioning was already observable (contra Faraoni, 2021). Second I will present articulatory data from an ongoing EMA study, which investigates the effect of lexical stress on the production of adjacent [ki] sequences. I will discuss the implications of these results for the μ–gesture model (Saltzmann, 2008; Katsika & Tsai, 2021) and for the diachronic emergence of stress-conditioning. Finally, I present a study investigating the perceptual consequences of prosodic enhancement on the discrimination of consonantal contrast.

Speaker: Giovanni Roversi (MIT)
Title: Adjectival “concord” in North Sámi is not concord (and it’s two different phenomena)
Time: Tuesday, April 18th, 1pm - 2pm

Abstract: Adjectives in North Sámi only inflect for case and number when they are predicative and
in case of NP ellipsis (NPE), and carry invariant morphology otherwise. I argue that the two contexts where they are inflected need two different accounts, and that neither of them reflects genuine nominal concord. In the NPE case, the morphology showing up on the adjective is the result of a stranded affix configuration caused by eliding the nP. I show that predicative adjectives, although they’re inflected, cannot be reduced to a case of NPE, but must get case/number features from a different source. I also propose an account of the adjectival morphology both when inflected and when not inflected.
— This is a practice talk for CLS, so I aim to talk for 20 minutes.

Speaker: Omri Doron (MIT)
Title: Hebrew nominal sentences wear their reconstruction on their sleeve
Time: Wednesday, April 19th, 1pm - 2pm

Abstract: Hebrew nominal sentences (with NP/AP as a predicate) sometimes contain what looks like a pronoun between the subject and the predicate (“Pron”), which agrees with the subject:

(1) dana (hi) gvoa
Dana (Pron.3FSG) tall
“Dana is tall”

Doron (1983) analyzes Pron as the realization of agreement features in I⁰, which is still considered the standard analysis of Pron. I point out that this analysis is unable to account for Pron’s complicated distribution and interpretative effects, and argue for an alternative analysis of Pron as a resumptive pronoun. I then show that Pron’s syntactic properties can be used to test different hypotheses in the literature regarding reconstruction effects.

Speaker: Janek Guerrini (Institut Jean Nicod, ENS)
Title: Revisiting kind predication
Time: Thursday, April 20th, 12:30pm - 2pm

Abstract: It is well-known that referential plurals support both collective predication with predicates of pluralities (‘the students are numerous’) and distributive predication with predicates of individuals (‘the students are blond’). Kind-referring plurals also support ‘collective’ kind predication, as in ‘birds are widespread’ (Carlson, 1977). However, with predicates of individuals, as in ‘birds fly’, it is usually assumed that there is no direct, distributive application of the predicate to the kind, but instead generic quantification on members of the kind (Krifka et al, 1995; Chierchia, 1998 a.o.). In this work, I argue that in sentences with predicates of individuals, kind-denoting plurals actually give rise to both logical forms: one that features generic quantification, and one that involves distributing a property directly over members of the kind. I then show how this insight resolves a number of long-standing puzzles in genericity.

Our recent MIT Indigenous Languages Initiative (MITILI) alum, Devon Denny (S.M., 2022) has accepted an offer to join UC San Diego’s Linguistics PhD program! He will be jointly supervised by Michelle Yuan (PhD, 2018) and Emily Clem. Devon plans to continue working on topics in the syntax of his native language, Navajo. Congratulations to UC San Diego and congratulations to Devon—-we’re so proud of you!

Speaker: Stan Zompì (MIT)
Title: *ABA in multidimensional paradigms
Time: Thursday, April 13th, 5pm - 6:30pm

Abstract:

The last decade and a half has witnessed a lot of research about *ABA universals—generalizations like “If a nominative and the corresponding dative have the same exponent, then the corresponding accusative has that exponent too” (Caha 2009; Smith et al. 2019). Most of the work on these universals has only focused on one ‘paradigm column’ at a time, by checking a given paradigm’s NOM.SG, ACC.SG, and DAT.SG, for example, with no heed to whether any of the exponents under scrutiny would also show up in that paradigm’s NOM.PL, ACC.PL, or DAT.PL. Recent literature, however, has pointed out that inspecting full paradigms is crucial to our understanding of *ABA, especially because some classic accounts that derive *ABA column-internally turn out to also make predictions about what may or may not happen across columns, and those predictions often appear to be wrong. In the domain of case, two proposals have recently been advanced to try and fix the problem: Christopoulos & Zompì’s (2022) and Caha’s (2023). I’ll argue that the former undergenerates (as Caha points out), whereas the latter overgenerates and, in so doing, misses a generalization. The challenge is to develop a theory whose power falls halfway between the two. To this end, I will explore the idea that exponents may both be underspecified and be overspecified with respect to their exponenda, and that each of these departures from a perfect match is penalized without necessarily being fatal—an intuition I will implement optimality-theoretically in terms of violable MAX and DEP constraints. I will argue that this derives the desired generalizations, and discuss some theoretical choice points and the (scant) evidence that might bear on them.

Speaker: Mitya Privoznov (MIT)
Title: CAUSE and causation in verb semantics. A modal account
Time: Wednesday, April 12th, 1pm - 2pm

Abstract:

In this talk, I will discuss the causation component in the lexical semantics of verb-like elements. More specifically, I will talk about four types of lexical items: (1) actuality entailing modals (see, e.g., Nadathur 2019, 2020); (2) verbs ‘make’ and ‘let’ (see, e.g., Raffy 2021); (3) morphological causatives (see, e.g., Lyutikova et al. 2006); (4) accomplishments (see, e.g., Dowty 1979, Kratzer 1996, Pylkkanen 2002, Ramchand 2008, Tatevosov 2008). All four of them have been argued to involve a causation component in their lexical meaning. Usually, causation is analyzed either as a relation between propositions (e.g., Dowty 1979) or events (e.g., Davidson 1967, Parsons 1990, Harley 1996, Pylkkanen 2002, Ramchand 2008) or as a force, which constitutes a separate ontological entity (e.g., Copley and Wolff 2014, Copley and Harley 2015, Raffy 2021). In the talk, I will try to argue for a modal account of the causation component in the lexical semantics of these items. According to this account, a causal operator is a modal quantifier with variable force (universal or existential) and fixed flavor. It is anchored to a situation (in terms of Kratzer 2007) and its domain of quantification includes all the causal chains initiated by the counterparts of the anchor situation in different possible worlds. I will try to show that this account adequately accounts for (1) the entailment relations between different causative predicates (e.g., ‘make’ and ‘let’, different interpretations of causatives); (2) underspecification of the causing situation (Agent vs. Causer theta-role, the causation by omission reading); and finally (3) the distribution of (anti)-actuality and (anti)-culmination entailments across aspectual contexts and negation.

Time: Friday, April 14th, 3.30-5pm 
Speakers: Jimmy Qiu, Vaishnavy Puvipalan & Lyn Tieu (University of Toronto)
Location: The talk will be virtual (on Zoom) but we will still be meeting in 32-D831!

Title: An experimental investigation of the inferences of emoji
Abstract:

Schlenker (2018) proposes a typology of ‘co-speech’, ‘pro-speech’, and ‘post-speech’ gestures, distinguished by whether the gestures are ‘external’ (can be eliminated without affecting the acceptability of the sentence) or ‘internal’, and whether the gestures occupy their own timeslot. Pierini (2021) extends Schlenker’s typology to emoji, identifying a set of predictions for how emoji should be interpreted in different kinds of sentences. First, ‘co-text emoji’, which immediately follow written text (e.g., John didn’t train today 🏋️) trigger ‘cosuppositional’ inferences (conditionalized assertion-dependent presuppositions) that project from embedded environments (if John had trained today, weight-lifting would have been involved). Second, ‘pro-text emoji’, which fully replace words (e.g., The egg will not 🐣), have an at-issue semantics and can trigger standard presuppositions (the egg is currently unhatched). Finally, ‘post-text emoji’, separated from accompanying text by a pause (e.g., appearing in the subsequent text message), generate supplements, which are degraded in negative environments. In this talk, we will describe a set of studies that investigate and ultimately provide experimental support for the predictions made in Pierini (2021).

Speaker: Kevin Ryan (Harvard University)
Title: On quantitative clash and lapse: the Old Tamil taḷai system of metrical linkage
Time: Tuesday, April 11th, 5pm - 6:30pm

Abstract:

Quantitative meters are familiarly described using (more or less flexible) templates which specify which kinds of metrical feet — and thus which sequences of heavy and light syllables — are permitted in which parts of the line. Such templates reflect recurring (ar)rhythmic structures. The oldest Old Tamil meters (early first millennium) work rather differently, in that any kind of metrical foot (within certain size limits) can appear anywhere in the line. What makes each meter distinctive is primarily its taḷai “linkage,” i.e. syntagmatic constraints on feet (e.g. foot type X can occur anywhere, but wherever it occurs, it cannot be followed by foot type Y). This system, which cannot be reduced to a fixed underlying pattern of prominence, is apparently indigenously Dravidian, making it one of the world’s few independent origins of quantitative meter. I analyze taḷai in terms of quantitative clash and lapse (along the lines of *Heavy-Heavy and *Light-Light; cf. Steriade 2017), arguing that standard (prominence-based) clash and lapse are irrelevant. Taḷai is further important for generative metrics in that (1) while some meters avoid clash and lapse, others prefer it (contra language, with music); (2) taḷai operates more stringently across feet than within them (contra a claimed universal); and (3) taḷai operates on the metrical parse rather than on intrinsic linguistic weight (contra strong prosodic metrics).

Speaker: Will Oxford (MIT, University of Manitoba)
Title: Contrastive, obligatory, and spurious voice
Time: Thursday, April 13th, 12:30pm - 2pm

Abstract:

Algonquian languages have a system of direct-inverse marking that is conditioned by a person hierarchy: 1/2 > 3 > 3′ > 3″ (where 3′ and 3″ represent obviative and “further obviative” third persons). Since a multi-level hierarchy such as this cannot be captured by a single feature, the apparent need to account for such hierarchies has led to some creative proposals about what syntax can do. I will argue that the Algonquian person hierarchy is in fact an illusion created by the use of marked voice morphology under three different conditions, which I refer to as “contrastive voice”, “obligatory voice”, and “spurious voice”. Each condition is responsible for one of the three rankings that make up the apparent hierarchy: the 3′ > 3″ ranking reflects contrastive voice, the 3 > 3′ ranking reflects obligatory voice, and the 1/2 > 3 ranking reflects spurious voice. I will show how this dissolution of the hierarchy improves our understanding of the data and explore its implications for formal models of voice and agreement.

Speaker: Ksenia Bogomolets (University of Auckland)
Title: How to get rid of idiosyncrasy in lexical accent systems
Time: Friday, April 14th, 3:30pm - 5pm

Abstract:

Phonological theory usually draws a broad distinction between two types of (stress-)accent systems: phonologically predictable accent vs. lexical accent systems. Lexical accent languages are often assumed to employ various complex morphology-based strategies to deal with accent assignment and competition (e.g. Alderete 1999; Bogomolets 2020; Hayes 1995; Revithiadou 1999; van der Hulst 2014). Such complexity is then taken to warrant idiosyncratic rules/constraints, which makes lexical accent systems look quite irregular in comparison to the phonologically predictable accent systems. Considering data from a Uto-Aztecan language Choguita Rarámuri, I argue that the surface complexity of lexical accent systems falls out from cross-linguistically attested morpho-syntactic configurations, while the accent assigning systems themselves only require simple, predominantly phonology-driven rules. These rules crucially are active cross-linguistically as well and are highly comparable to those found in phonologically predictable accent systems.

Speaker: Zachary Satoshi Feldcamp (MIT)
Title: Predicate inversion as A-movement
Time: Tuesday, April 4th, 1pm - 2pm

Abstract: Predicate inversion (PI) constructions (1) involve a marked order of constituents in which the predicate (underlined) appears to occupy the typical derived subject position, and the would-be subject (italicized), which often receives the nuclear pitch accent and focus, occurs to the right of the auxiliary be (Emonds 1970; 1976; Birner & Ward 1998; Samko 2016; Thoms & Walkden 2019).

(1) Performing at the concert next month will be SAM.

The absence of weak cross-over effects and of reconstruction for Condition C suggests that the predicate undergoes A-movement, which is initially puzzling, given what is known about A-movement, locality, and the interpretation of moved predicates, additionally raising questions about the possible role of information-structure in syntactic derivations. Contra Samko (2014; 2016), I argue that PI is not driven by information-structural features, because it is not subject to any firm information-structural conditions.

I propose that PI is the result of T agreeing with the predicate instead of the would-be subject, possible whenever the predicate has (incomplete) φ-features valued via concord with the would-be subject. No locality issues arise, because both subject and predicate may move in a nesting configuration to specifiers of Aux be, which position can be directly observed in passive expletive constructions (Rezac 2006). Given the cross-linguistic generalization that non-nominal lexical categories do not bear person features, the analysis predicts that T, which has agreed with the predicate, lacks person in PI. I show that this prediction is borne out. Despite its unusual appearance, then, PI corroborates the hypothesis that A-movement is the reflex of φ-Agree (van Urk 2015).

Speaker: Mirella L. Blum (University of Edinburgh)
Title: Correspondence and variation in Dinka tone
Time: Monday, April 3rd, 5pm - 6:30pm

Abstract: This talk examines tonal phonology across dialects of the Dinka language (West Nilotic, South Sudan). Dinka has a highly complex sound system, with a three-level vowel length contrast, a voice quality contrast, and tone; all elements are both lexical and grammatical, and most of the language’s morphology is expressed through its suprasegmentals. Along with the high functional load of tone and extensive tone sandhi, the varying number of tones— dialects of Dinka have either three or four tones—has led to the impression that the tone systems of different varieties of Dinka cannot be compared, and that systematic correspondences do not exist. In contrast, I show how tones correspond across dialects of Dinka, depending in part on the vowel grade system—alternations of vowel quality and length pervasive throughout numerous areas of Dinka morphology. I show that dialects likely shift from four to three tones, not the other way around, and I discuss the range of tonal processes across dialects of Dinka, how the processes relate to one another, and how they hint at the evolution of the tone systems of individual dialects.

Speaker: Mitya Privoznov (MIT)
Title: The syntax of presupposition projection
Time: Thursday, April 6th, 12:30pm - 2pm

Abstract: Consider the following pair of sentences:

(1) Context: I don’t know whether Rosa ever smoked or not.
a. #But I don’t think that [she stopped smoking altogether] and [she used to smoke Belomor].
b. But I don’t think that [she used to smoke Belomor] and [stopped smoking altogether].

The sentence in (1a) sounds infelicitous in the given context, while the sentence in (1b) does not, see Mandelkern et al. (2020), Kalomoiros (2021) and Kalomoiros and Schwarz (2021) for experimental evidence supporting this claim. The sentence in (1a) contains a presupposition trigger stopped, which introduces a presupposition p = ‘she used to smoke’ at the level of the clause she stopped smoking altogether. Being a presupposition, p projects to the level of the whole sentence in (1a) from under and, think and negation in the main clause. It comes into a contradiction with the given context (I don’t know whether Rosa ever smoked or not) and the sentence is predictably judged as infelicitous. The sentence in (1b) is truth-conditionally equivalent to (1a). Moreover, it contains the same presupposition trigger stopped, which introduces the same presupposition, which should project to the level of the whole sentence in (1b) and come into a contradiction with the given context, so (1b) should also be judged infelicitous, but it is not.

Contrasts like the one in (1) have led many researchers to the conclusion that a presupposition does not impose a requirement on the global context of the whole sentence that contains the trigger, but rather on the local context of the trigger. The local context is calculated based on the global context and the syntactic context of the trigger (other material in the sentence). In this talk, I will consider various ways of calculating the syntactic context of the trigger and show that the algorithms that have been proposed in the literature (a lexical one, a linear one and a quantifier-type one) all have unsatisfactory consequences. As a response to these problems, I will propose a syntactic algorithm, which avoids those consequences and leads to an unexpected relation between presupposition projection and the nature of syntactic derivation and the nature of Spell Out / Transfer.

Speaker: Ofelia Zepeda (The University of Arizona)
Title: The Varied Roles of Native American/Indigenous Linguists
Time: Friday, April 7th, 3:30pm - 5pm

Abstract: In the late 20th Century, a small handful of Native Americans who were speakers of their language were taken under the wing by a few established non-Native linguists with the intent to introduce them to linguistics and train them in the field so that they might become linguists working on their own language or related fields. This presentation will provide a review of the modern and short history of these Native American/Indigenous linguists. The presentation will then elaborate on the newest and upcoming group of Native American/Indigenous linguists and how some of them were called, designate, propelled or chose to go into linguistics primarily due to the dramatic and devastating impact of language endangerment and language loss in their communities. Having worked with some of these students I find that they know that they can find themselves in seemingly insurmountable situations in this field and within their communities. These students persevere understanding that skills in linguistics is one of the important tools in their efforts in language reclamation.

Speaker: Juan Cancel (MIT)
Title: The suffix -γi in Central Alaskan Yupik: Antipassive and Applicative Marker, or just an Applicative Marker
Time: Tuesday, March 21st, 1pm - 2pm

Abstract: For my undergrad thesis, I looked at -γi, a valency-changing suffix in Central Alaskan Yupik (CAY) with adversative/malefactive qualities. When the suffix appears in ergative-aligned transitive verbs, -γi is the most productive marker of the antipassive, but when it appears in accusative-aligned transitive verbs, -γi marks the applicative, adding a new argument with absolutive case. In addition to this, when it appears in intransitive verbs, -γi still marks the applicative, but the newly added argument appears with ergative case. Although most of the literature has subsumed -γi as functioning both as an antipassive and an applicative marker, I argued that -γi is just an applicative marker and that its antipassive use is the result of zero -derivation, another valency-changing operation in CAY.

Speaker: Tanya Bondarenko (Harvard Universiy)
Title: Veridicality mismatches in Javanese
Time: Wednesday, March 22nd, 1pm - 2pm

Abstract: In this talk I discuss a violation of the Spector & Egré’s generalization about the correlation between veridicality in declarative and interrogative embeddings, (1).

(1)Spector & Egre (2015)’s Generalization:
A responsive predicate is veridical with respect to its interrogative complement (like know + question = knowing the true answer to the question) if and only if it is veridical with respect to its declarative complements as well (know + declarative entails —in fact presupposes— that the declarative is true). (Spector & Egre 2015: 1732)

The counterexample to the generalization comes from Javanese: the verbs `know’ (ngêrti) and `remember’ (kelingan) in this language are veridical with respect to question embedding, (3), but, surprisingly, are not veridical with respect to the declarative embedding, (2).

I propose that this pattern emerges from a combination of two factors: (i) embedded clauses in Javanese are always adjuncts, and compose with verbs as modi ers that specify the propositional content of the verbal eventuality (see Bogal-Allbritten 2016, Elliott 2017, Bochnak & Hanink 2021, Bondarenko 2022, a.o.), which leads to the lack of veridicality with declarative embedding; (ii) propositional content associated with ngêrti `know’ and kelingan `remember’ is required to be a single proposition, which in cases of interrogative embedding triggers insertion of an answerhood operator (Dayal 1996) that gives rise to veridicality. I furthermore show that while ngêrti `know’ is non-veridical with declarative complements, it is different from `think/believe’: it comes with a presupposition that the embedded proposition is not known to be false. I suggest to encode this restriction as a definedness condition on the propositional content associated with the mental state that ngêrti denotes.

References
Bogal-Allbritten, Elizabeth. 2016. Building meaning in Navajo. University of Massachusetts Amherst dissertation.
Bochnak, Ryan & Emily Hanink. 2021. Clausal embedding in Washo: Complementation vs. modi cation. Natural Language & Linguistic Theory. 1-44.
Bondarenko, Tatiana. 2022. Anatomy of An Attitude. MIT dissertation.
Dayal, Veneeta. 1996. Locality in WH-quanti cation: Questions and relative clauses in Hindi. Dordrecht: Kluwer Academic Publishers.
Elliott, Patrick. 2017. Elements of clausal embedding. UCL (University College London) dissertation.
Spector, B. and Egr e, P., 2015. A uniform semantics for embedded interrogatives: An answer, not necessarily the answer. Synthese, 192, pp.1729-1784.

Students in 24.S95 Linguistics in K-12 Education—Aspen Abner, Cora Lesure, Lorenzo Pinton, Christopher Legerme, and Vincent Zu—are all smiles after teaching 7th and 8th graders at MIT’s Spring Spark on Saturday, March 18. Not surprisingly, the course that they designed and taught, Linguistics: The Science of Language, sparked a lot of interest!

Speaker: Ksenia Ershova (MIT)
Title: Licensed to license: Deficient probes in West Circassian nominalizations
Time: Thursday, March 23rd, 12:30pm - 2pm

Abstract: This talk argues for a context-dependent, counter-cyclic view of φ-probe feature deficiency: verbal probes must be licensed by the highest head in the extended projection (C^0) in order to agree with and license nominal arguments. This is supported by verbal nominalizations in West Circassian (WC), which, despite being as large as TP and including the full range of verbal φ-probes, cannot expone verbal φ-agreement, assign case or license arguments. The proposal may be extended to other instances of probes being licensed by a higher projection, e.g. ERG-assigning v^0 in Hindi (Legate 2008) or genitive of negation in Russian (Bailyn 2004).

Speaker: Vera Hohaus (MIT)
Title: Linguistic Illusions Revisited: The Role of Maximal Informativity
Time: Friday, March 24th, 3:30pm - 5pm

Abstract: Building on joint work with Nadine Bade (Universität Potsdam) and Ryan Walter Smith (The University of Manchester), this talk revisits two prominent cases of linguistic illusions that involve comparison, so-called depth-charge sentences like (1) and under-over illusions like (2).

(1) No head injury is too insignificant to be ignored.

(2) The importance of the rainforest cannot be underestimated.

We demonstrate that both the pedantic and the intended interpretation of these sentences can be derived compositionally. The alleged illusion arises from the interaction of maximal informativity with the monotonicity of the degree predicate underlying the comparison.

The 47th Penn Linguistics Conference took place at the University of Pennsylvania on over the weekend (March 18-19). The program included presentations by the following members of the MIT community:

• Adèle Hénot-Mortier (4th year): Superiority effects and the French Plural Pronoun Construction
• Katya Morgunova and Anastasia Tsilia (2nd year): Why would you D that? On the D-layer in Greek clausal subjects

The 5^th Crete Summer School of Linguistics, or ‘CreteLing’, will take place from July 15 to July 28, 2023 at the University of Crete in Rethymno, Greece.

As in other years, a number of current faculty, as well as current and former students will be involved.

Current MIT faculty Adam Albright, Kai von Fintel, Shigeru Miyagawa, David Pesetsky, Norvin Richards, and Donca Steriade will be teaching classes, along with alumni Karlos Arregi, Paul Kiparsky, Ömer Demirok, Pritty Patel-Grosz, Doug Pulleyblank, and Philippe Schlenker. They will be joined by many wonderful colleagues from around the world.

Besides the wide range of courses offered across four parallel sessions, the summer school will feature two workshops: ‘Workshop on Language Acquisition’, organized by Artemis Alexiadou, Maria Teresa Guasti, and Uli Sauerland, and ‘Gender Markedness and Defaults’, organized by Luke Adamson. 

For more information, as well as the application form, please consult the school’s website: https://linguistics.philology.uoc.gr/cssl23/

Speaker: Ka-Fai Yip (Yale University)
Title: A compositional account of “only” doubling
Time: Wednesday, March 15th, 1pm - 2pm

Abstract: Cross-linguistically, exclusive particles ‘only’ may be doubled with a single focus association, posing a problem for the Principle of Compositionality (Dutch: Barbiers 2014; German: Hole 2015; Korean: Lee 2005; Mandarin: Hole 2017, Sun 2021; Vietnamese: Hole 2017, Erlewine 2017; i.a.). The predominant account, operator-particle approach (Quek & Hirsch’s 2017; cf. Branan & Erlewine 2023), explains doubling of adverbial and adfocus particles by treating the latter as semantically vacuous concord markers that establish a syntactic dependency with the former. In this study, I focus on an understudied case of ‘only’ doubling of adverbial particles and sentence-final particles (SFPs) in Cantonese (Law 2004, Lee 2019), which is also attested in Mandarin (Erlewine 2011) and Vietnamese (Hole 2013). While I follow the tenet of the operator-particle approach that one particle is dependent on another one (which is an operator), I pursue a different route in two-dimensional semantics concerning at-issueness and argue that both particles have focus-sensitive contributions, distributed in different meaning dimensions. Specifically, the SFPs relate the focus alternative set quantified by adverbial ‘only’, the genuine exclusive operator, to the discourse by requiring the excluded alternatives to be contextually salient. Informally, the SFPs add a “contrastive” flavor to the exclusive focus. I further demonstrate how the dependency between the two particles may be accounted for by the semantics of the SFPs in terms of distinguishing excluded alternatives from the presupposed proposition, and offer a compositional analysis. I will also address issues raised by scalar readings and how to unify non-scalar and scalar uses in ‘only’ doubling.

This week, Shane Steinert-Threlkeld (University of Washington) will give a colloquium talk and two-part minicourse as the first invited speaker of the Breakstone Speaker Series on Language, Mind, and Computation. This brand new speaker series is an interdisciplinary collaboration between faculty from MIT Linguistics, MIT Brain and Cognitive Sciences, and MIT CSAIL. The funding for the series comes in the form of a donation by Micha Breakstone (PhD, Hebrew University of Jerusalem). 

Abstracts and details about Shane’s visit can be found below. 

Colloquium: Unnatural Language Semantics

Mini-course: The Artificial Language Toolkit

Speaker: Keny Chatain (Institut Jean Nicod (ENS))
Title: Reducing Pronoun Accessibility To Presupposition Satisfaction
Time: Thursday, March 9th, 12:30pm - 2pm

Abstract: A pronoun cannot always co-refer with a given DP: while the cases in (1a-c) are natural, the cases in (1d-e) are not interpretable. The problem of pronoun accessibility is the problem of determining which antecedent-pronoun configuration are licit, which ones give rise to deviance. The cases in (1) point to a simple generalization: the pronoun “it” can co-refer with “a phone-book” if and only if the existence of a phone-book can be taken for granted at the point where the pronoun “it” is used. Simple though it may seem, this generalization is not fully validated by many current theories of pronouns. Such theories typically under-generate, failing to license cases like (1c) as well as other more complicated examples.

(1)
a. There is a phone-book₇ and it₇ is in the cabinet.
b. If there is a phone-book₇, it₇ is in the cabinet.
c. Either there isn’t a phone-book₇ and it₇ is in the cabinet. (attributed to Partee)
d. # Either there is a phone-book₇ or it₇ is in the cabinet.
e. # There might be a phone-book₇ and it₇ is in the cabinet.

Taking the generalization at face-value, I propose a system where a pronoun can be interpreted if and only if the existence of a witness - a phone-book in the cases above - can be presupposed. This theory builds on insights from E-type theories (Evans, 1980 ; Heim, 1990 ; Elbourne, 2005) but drops some of the assumptions that have made such theories inviable, like uniqueness.

The benefits are conceptual and empirical. Conceptually, by reducing pronoun accessibility to presupposition satisfaction, the proposal can build upon so-called “explanatory” theories of presuppositions (Schlenker, 2009 ; George, 2008 ; Fox, 2013, a.o.). Such theories derive discourse effects from truth-conditional meaning, instead of baking these effects into meanings themselves (Soames, 1989), as in Dynamic Semantics (Heim, 1983, a. o.). Second, the proposal has a broad empirical coverage: it explains the original cases in (1), but also the more complex quantified cases in (2) of quantifier subordination (Roberts, 1987) and donkey anaphora (Geach, 1962). In addition, it makes a range of new predictions, regarding the possibility of cataphora and pronouns licensed by pragmatic inferences.

(2)
a. Every farmer who has a donkey feeds it hay.
b. Every farmer has a donkey. Few of them feed it hay.

Speaker: Vera Gribanova (Stanford University)
Title: On the sources of case (dis-)connectivity in two types of Russian TP ellipsis
Time: Friday, March 10th, 3:30pm - 5pm

Abstract: Discussions of the identity relation in constituent ellipsis licensing often take for granted, either explicitly or implicitly, the idea that the identity relation in ellipsis ought to be uniform, and applicable across different ellipsis configurations and languages. Recent investigations of this relation — Rudin 2019, Kroll 2019, Anand, Hardt, and McCloskey 2022, Stigliano 2022 — have provided novel evidence and arguments in support of the view that the domain of application of the identity relation is not always coextensive with the domain of the ellipsis itself. For example, although the prevailing view of English sluicing as TP ellipsis historically took the domain of the identity relation to likewise be the TP (Merchant 2001), one of the main findings of the UCSC sluicing dataset is that material above the level of vP — e.g. tense/finiteness, modality, and polarity — can undergo felicitous mismatches. Generalizing beyond English sluicing, this raises the question of whether the domain of ellipsis identity must be a proper subset of the domain of ellipsis itself, or if the specific size of the domain relevant for the identity relation may be variable across languages and ellipsis configurations.

In this talk, I present an investigation of some asymmetries in how case connectivity is enforced in two types of Russian clausal (TP) ellipsis — contrastive polarity ellipsis and fragment answers — and develop an analysis explaining why these asymmetries take the shape that they do. The case study leverages the availability of a well-known case alternation between structural (nominative/accusative) case and genitive case under negation. The first asymmetry is that case connectivity on remnants of these two ellipsis types is enforced fully only in fragment answers, but not in contrastive polarity ellipsis, in which a contrastive DP is fronted to the left periphery, preceding a polar particle (‘yes’ or ‘no’). The second asymmetry is that in contrastive polarity ellipsis, genitive patients under negation in the antecedent can correspond to an accusative patient remnant outside the ellipsis site, but not the reverse. To capture these asymmetries, I develop an analysis of the system of licensing relations that connects the syntax of polarity expression, negative concord, and genitive of negation, and combine this with a formulation of the identity relation in ellipsis in which head-to-head identity between the elided material and the antecedent must be invoked (Saab 2008, 2010, 2022, Tanaka 2011, Rudin 2019, Stigliano 2022).

For the asymmetries between these two types of Russian TP ellipsis to emerge within an internally consistent system of analytical commitments, it is critical that the domain of evaluation for identity be larger than in English sluicing, and likely coextensive with the elided TP. This finding supports a view in which the domain of evaluation for the identity relation in ellipsis may vary across languages and ellipsis types. In the latter part of the talk, I point to a view of ellipsis licensing that can straightforwardly capture such variation, and which arises directly from unifying existing analyses (Aelbrecht 2010, Stigliano 2022) in which certain sub-parts of the ellipsis function — non-pronunciation, syntactic licensing, and the identity relation — can be either grouped together, or broken up across several distinct heads in the clausal spine.

References

Aelbrecht, Lobke. 2010. The syntactic licensing of ellipsis. John Benjamins. Anand, Pranav, Daniel Hardt, and James McCloskey. To appear. The domain of formal matching in sluicing. Linguistic Inquiry. Kroll. Margaret. 2019. Polarity reversals under sluicing. Semantics and Pragmatics 12. Merchant, Jason. 2001. The Syntax of Silence: Sluicing, Islands, and the Theory of Ellipsis. Oxford: Oxford University Press. Ross, John Robert. 1969. Guess who? In Proceedings of the 5th Annual Meeting of the Chicago Linguistic Society. Robert I. Binnick, Alice Davison, Georgia M. Green, and Jerry L. Morgan (eds.), 252–286. Rudin, Deniz. 2019. Head-based syntactic identity in sluicing. Linguistic Inquiry 50 (2): 253–283. Tanaka, Hidekazu. 2011. Syntactic identity and ellipsis. The Linguistic Review 28: 79–110. Saab, Andrés. 2008. Hacia una teoría de la identidad parcial en las elipsis. PhD Dissertation, Universidad de Buenos Aires. Saab, Andrés. 2010. Silent interactions: Spanish TP-ellipsis and the theory of island repair. Probus 22 (1): 73–116. Saab, Andrés. 2022. Grammatical silences from syntax to morphology. A model for the timing of ellipsis. In The Derivational Timing of Ellipsis, ed. Güliz Günes and Anikó Lipták. Oxford: Oxford University Press. 
Stigliano, Laura. 2022. The silence of syntax: A theory of ellipsis licensing and identity. PhD dissertation, University of Chicago.

On Thursday, March 2, 1pm, MIT Indigenous Languages Initiative (MITILI) alum Annauk Olin (S.M. 2021) will lead the first Inupiaq working group meeting. On Thursday and at future meetings, the group will aim to piece together grammatical aspects of the Inupiaq language, specific to the Native Village of Shishmaref. All are welcome to join: for more information, please contact Annauk (annauk@mit.edu).  

Speaker: Boer Fu (MIT)
Title: Mandarin Glide Segmentation: A Language Game Experiment
Time: Monday, February 27th, 5pm - 6:30pm

Abstract: The acquisition of a phonological grammar requires the segmentation of audio input into individual consonants and vowels as a first step. Segmentation is often taken as a given in the study of alphabetical languages, but triggers much debate in Mandarin phonology, especially around the prenuclear glide G in the syllable template CGVX (e.g. [j] in [ɕjɑŋ] ‘fragrance’). Some scholars treat the glide as an independent segment (e.g. Lin 1989). Others claim it is a secondary articulation of the onset (e.g. Duanmu 2007). Still others argue it is nothing more than a natural CV transition (e.g. Ladefoged & Maddieson 1996). These proposals for glide segmentation assume uniform segmentation among Mandarin speakers. I present a language game experiment that shows there is much speaker variation in glide segmentation. The language game is based on Chinese fanqie secret languages (see Chao 1931). In the experiment, Mandarin speakers are asked to swap the onsets of a disyllabic word, in order to encode it as a secret message. For example, [kʰa fej] ‘coffee’ is encoded as [fa kʰej]. What a speaker chooses to do with the glide can inform us of their segmentation of the sound.

Speaker: Giovanni Roversi (MIT)
Title: The morphology of North Sámi adjectives
Time: Thursday, February 23rd, 5pm - 6:30pm

Abstract: North Sámi adjectives are sometimes inflected for case and number, and sometimes not. The pattern is quite complex, and reminiscent of other languages (Hungarian, Turkish, etc.) for which analyses exist on the market, but with added complications that the existing analyses may or may not capture without patches. I want to present some work in progress that tries to take a stab at developing a full account of the North Sámi data, and I’ll highlight the parts that still need some ironing out.

Speaker: Amir Anvari (MIT)
Title: Origins of Conservativity
Time: Wednesday, February 22nd, 1pm - 2pm

Abstract: While that natural language determiners denote conservative functions (Keenan & Stavi 1986) is perhaps the most commented on semantic universal, it nevertheless remains unexplained. I will revisit the most promising approach to this puzzle based on the idea that copy theory of movement (Chomsky 1995) makes non-conservative determiners semantically vacuous (Chierchia 1995, Fox 2002,2003, Sportiche 2005). The only serious study of this idea was done by Romoli (2015) with the surprising conclusion that non-conservative determiners may very exist but we cannot be sure as, within copy theory of movement, such determiners are bound to yield the same truth-conditions as conservative ones. First, I will address Romoli’s skepticism by leveraging presupposition projection to argue that all determiners are in fact lexically conservative. Second, I will argue that, with appropriate auxiliary assumptions about the denotation of determiners, Schlenker’s (2009) theory of presupposition projection coupled with Fox’s (2002,2003) rule of trace conversion correctly predicts all determiners to be conservative within the copy-theoretic framework. Third, I will explore a double strengthening of this result by suggesting that (a) conservativity is explained by presupposition projection even without copy theory of movement and (b) the fact that natural language determiners denote generalized (i.e. restricted) quantifiers to begin with is also due to presupposition projection. Finally, I will argue that this work raises the bar on what counts as an acceptable, explanatory theory of presupposition projection by comparing the main result based on Schlenker’s theory of presupposition projection with the predictions made by the Strong Kleene system (Fox 2012).

The Workshop on Altaic Formal Linguistics (WAFL) 17 will be hosted by the National University of Mongolia, Ulaanbaatar, September 27, 28, 29, 2023. This year, our colleague Norvin Richards is among the invited speakers!

Abstracts are invited for 20-minute talks (plus 10-minute discussions) and for posters on topics dealing with formal aspects of any area of theoretical Altaic linguistics, including phonology, morphology, syntax, semantics, or pragmatics. The term ‘Altaic’ is understood to include Turkic, Mongolic, and Tungusic languages, as well as Korean, Japanese, Ryukyuan, and Ainu.

The deadline for abstracts is April 14, 2023.

Speaker: Amir Anvari (MIT)
Title: Origins of Conservativity
Time: Wednesday, February 15th, 1pm - 2pm

Abstract: While that natural language determiners denote conservative functions (Keenan & Stavi 1986) is perhaps the most commented on semantic universal, it nevertheless remains unexplained. I will revisit the most promising approach to this puzzle based on the idea that copy theory of movement (Chomsky 1995) makes non-conservative determiners semantically vacuous (Chierchia 1995, Fox 2002,2003, Sportiche 2005). The only serious study of this idea was done by Romoli (2015) with the surprising conclusion that non-conservative determiners may very exist but we cannot be sure as, within copy theory of movement, such determiners are bound to yield the same truth-conditions as conservative ones. First, I will address Romoli’s skepticism by leveraging presupposition projection to argue that all determiners are in fact lexically conservative. Second, I will argue that, with appropriate auxiliary assumptions about the denotation of determiners, Schlenker’s (2009) theory of presupposition projection coupled with Fox’s (2002,2003) rule of trace conversion correctly predicts all determiners to be conservative within the copy-theoretic framework. Third, I will explore a double strengthening of this result by suggesting that (a) conservativity is explained by presupposition projection even without copy theory of movement and (b) the fact that natural language determiners denote generalized (i.e. restricted) quantifiers to begin with is also due to presupposition projection. Finally, I will argue that this work raises the bar on what counts as an acceptable, explanatory theory of presupposition projection by comparing the main result based on Schlenker’s theory of presupposition projection with the predictions made by the Strong Kleene system (Fox 2012).

Welcome to the first edition of Whamit! for Spring 2023! After our winter hiatus, Whamit! is back to regular weekly editions during the semester. 

Whamit! is the MIT Linguistics newsletter, published every Monday (Tuesday if Monday is a holiday). The editorial staff consists of Adam Albright, Kai von Fintel, David Pesetsky, and Keely New. 

To submit items for inclusion in Whamit!, please send an email to whamit@mit.edu by Sunday 6pm.

This week, we’re delighted to welcome Juliet Stanton in the deparment for an extended colloquium visit! As part of her visit, Juliet will teach a minicourse over two days.

Speaker: Juliet Stanton (New York University)

Minicourse title: Hiatus resolution in the Cangin languages

Days and times: Tuesday (2/7), 12.30-2pm; Wednesday (2/8), 1-2.30pm

Location: 32-D461

Abstract: The Cangin languages are a small subgroup of Atlantic languages spoken in Senegal. All ban hiatus (sequences of two consecutive vowels) but resolve it in different ways. In two languages, the resolution strategies are not surprising: Palor and Ndut resolve hiatus through a combination of vowel deletion and glide epenthesis. In the others, however, the resolution strategies are more surprising: Noon and Laalaa (and probably Saafi) resolve hiatus through epenthesis of the coronal nasal [n]. Most of the lectures will be devoted to summarizing the patterns, sketching analyses, and determining the degree to which the [n] epenthesis pattern is predicted (or not) by extant theories of consonant epenthesis. This is work in progress, so your thoughts will be welcome!

Speaker: Juliet Stanton (NYU)
Title: Phonetic rhythm in American English -ization
Time: Friday, February 10th, 3:30pm - 5pm

Abstract: It is commonly assumed that *Clash and *Lapse evaluate syllable-sized constituents: a sequence of two adjacent stressed syllables violates *Clash, while a sequence of two stressless syllables violates *Lapse (see e.g., Prince 1983, Gordon 2002 for *Clash; Nespor & Vogel 1989, Green & Kenstowicz 1995, Gordon 2002 for *Lapse). In this talk I argue, based on patterns of secondary stress in American English -ization, that there exist rhythmic constraints that are evaluated with respect to (normalized) duration. I present evidence consistent with this claim from corpus, judgment, and production studies.

Speaker: Danfeng Wu (University of Oxford)
Title: Syntax and prosody of coordination
Time: Thursday, February 9th, 12:30pm - 2pm

Abstract: In this talk I present a generalized syntactic analysis of coordination, with a focus on corrective but sentences (e.g., (1a-b)) as the case study. These sentences involve coordination by but, and require presence of negation in the first conjunct and absence of negation in the second (e.g., *Max eats spinach but chard; *Max doesn’t eat spinach but not chard).

(1) a. Max doesn’t eat spinach but chard. b. Max eats not spinach but chard.

Evidence based on scope suggests that negation has two positions in these sentences, though we only hear one. This analysis is consonant with previous proposals for focus-sensitive operators such as the Question-particle and only (e.g. Cable 2007; Hirsch 2017), suggesting that perhaps all focus-sensitive operators have two positions in a sentence. Then I argue using evidence involving constituency, scope and other phenomena that ellipsis can occur to obscure the positions of negation.

In addition to the syntactic-semantic arguments for ellipsis, I present evidence from a prosodic experiment, following a tradition in the literature that draws evidence for syntactic theories from prosodic evidence (e.g. Bresnan 1971; Clemens & Coon 2018). In this experiment, I show that this syntactic analysis makes predictions with measurable effects in prosodic phrasing. The experimental results also suggest that prosodic structure corresponds to syntactic structure more closely than some theories previously claimed.

Course announcements in this post:

• Topics in Syntax (24.956)
• Topics in Semantics (24.979)
• Topics in Computational Phonology (24.981)
• Topics in Computational Linguistics (24.982)
• Special topics: Linguistics in K-12 education (24.S95)

24.956: Topics in Syntax

• Instructors: Ksenia Ershova and Will Oxford
• Wednesdays, 10am-1pm
• Room: 32-D461

Subject is one of the most fundamental and most frequently appealed to notions in the discussion of argument asymmetries cross-linguistically. Subjects are taken to display a cluster of properties, which in tree-geometric terms are associated with being the structurally highest argument in the clause. Properties typically associated with subjects include: (i) unmarked (nominative) case; (ii) the ability to control verbal agreement; (iii) the ability to bind anaphors; (iv) the ability to be PRO and to participate in raising; (v) agentivity and thematic prominence; (vi) topicality; (vii) accessibility for wh-movement. In modern Minimalism these properties are distributed across several positions in the clause, but tend to converge on a single nominal due to standard constraints on locality and movement. In this seminar we will explore phenomena that challenge a universally homogeneous notion of subjecthood, focusing on cases where the subject displays only a subset of typical subjecthood properties, or where subjecthood properties are distributed across more than one argument in the clause. We will discuss both the empirical landscape of research on subjecthood and the implications that research has for syntactic theory and our understanding of locality, intervention, licensing, case, agreement, thematic and structural prominence, etc.

24.979: Topics in Semantics 

• Instructors: Amir Anvari and Kai von Fintel
• Mondays, 2-5pm
• Room: 32-D461

This semester, we will explore the philosophy of natural language semantics, or meta-semantics. The overarching question will be: what has to be the case for a prominent branch of formal semantics, often referred to as Heim-and-Kratzer semantics, and various specific proposal made within it over years, to make any sense? The course will be split in two parts. In part one, our attention will be focused on what makes formal semantics formal: the emphasis on entailment and contradiction. These concepts seem to play important and diverse roles in semantic theorizing and, by extension, linguistic theory. What does this tell us about language and, by extension, how the mind works? In part two, we will discuss a series of issues, some of which may already be raised in part one. Possible topics include internalism vs externalism; the prospects for referential semantics; the idea of natural language metaphysics/ontology; the position of semantics vis-a-vis cognitive science and/or philosophy; questions of expressive power and type economy; issues of modularity; the connection between language and thought; critiques of mainstream formal semantics from authors like Chomsky, Jackendoff, and Pietroski.

We concur with the following from Bob Stalnaker’s seminar description: “The schedule will be flexible and open ended, following the discussion where it leads, and spending as much time on each topic as it seems to need. That is, we will make it up as we go along.”

As usual, to receive credit, we expect active participation in seminar meetings, weekly emailed questions and comments, and a final paper.

24.981: Topics in Computational Phonology

• Instructor: Giorgio Magri
• Tuesdays, 2-5
• Room: 66-160

This year’s installment of 24.981 (Topics in Computational Phonology) will focus on categorical phonology and will address the perennial issue of OT’s strict domination versus HG’s constraint weighting and the additive (or cumulative or gang) effects precluded to the former but allowed by the latter. The tentative plan includes the following issues.

[1] Defining additive effects.
I will put forward a new, purely extensional definition of ‘additive’ effects. We will then discuss it through a couple of concrete examples
that have figured prominently in the literature on additive effects. If the definition makes sense, it affords a way to talk about additive effects (and thus of comparing OT and HG) that makes no reference to constraints, weights, or rankings.

[2] OT and HG as two sides of the same coin. I will introduce the OT and HG implementations of constraint-based
phonology from scratch, trying to formally deduce both of them from the same axiom on additive effects in one fell swoop. This suggests
that, when OT and HG are construed within the huge space of all logically possible implementations of constraint-based phonology, they are quite similar in terms of additive effects: neither of them yields many.

[3] Additive effects in OT and HG.
We will then focus on additive effects in HG and OT, trying to understand which additive effects are indeed within HG’s reach and
which instead require something like ‘constraint conjunction’ in both OT and HG.

[4] OT as ‘margin-free’ HG.
We will then switch to learnability, focusing on ‘online’ or ‘error-driven’ learners (that is, learners that discard each piece of data after having encountered it, rather than storing it). I will review the classical ERCD/GLA theory of online learning in OT. We will then discuss the ideas of ‘margin’ and ‘kernel trick’ in the context of HG. And I will try to conclude that HG, contrary to OT, comes with no good online learners.

[5] Strict domination and exponential update rules.
OT’s strict domination can be mimicked by HG weighting, as long as the weights decay exponentially fast (relative to the size of the constraints). Building on this observation, we will investigate whether learning algorithms with an exponential update rule can be rebooted as OT learning algorithms. We will focus on two cases: ‘AdaBoost’, a batch iterative algorithm; and Winnow, an online algorithm.

Quoting from Amir and Kai’s quote of Bob’s seminar description: “The schedule will be flexible and open ended, following the discussion
where it leads, and spending as much time on each topic as it seems to need. That is, we will make it up as we go along.” As for requirements, I propose the following three: (1) a squib to turn in at the end of the course; (2) four simple problem sets (one every two weeks, during the first two months of the course); (3) taking turns at transcribing classes (that will be taught out of a lean handout, mostly at the blackboard). This course can be used to satisfy the program’s acquisition requirement with a suitable choice of the topic for the final project. Please consult me at the beginning of the semester if you are planning to do so.

24.982: Topics in Computational Linguistics

• Instructor: Forrest Davis
• Tuesdays, 10am-1pm
• Room: 32-D461

We will be exploring the relationship between computational models and linguistic theory, with a particular focus on neural models of language (e.g., GPT-3). The main theme of the course will be how neural models should relate to a theory of language. As a means of orienting our initial discussions, we will focus on three papers:

1. Wilcox, Futrell, and Levy. (2022). Using Computational Models to Test Syntactic Learnability. Linguistic Inquiry.
2. Baroni. (2022). On the proper role of linguistically-oriented deep net analysis in linguistic theorizing. In Algebraic Structures in Natural Language.
3. Steinert-Threlkeld and Szymanik (2019). Learnability and semantic universals. Semantics and Pragmatics.

Taking these papers in turn, we will touch on a variety of topics ranging from the relationship (or lack thereof) between grammaticality and probability, acceptability judgments, learnability, and the poverty of the stimulus. Background material will be supplemented as needed and topics expanded depending on interest. Possible additional directions include, probing models for syntactic trees (e.g., Hewitt and Manning, 2019, A Structural Probe for Finding Syntax in Word Representations), superficialism (e.g., Rey,, 2020, Representation of Language), and/or meta-learning for adding linguistic knowledge to models (e.g., McCoy et al., 2020, Universal linguistic inductive biases via meta-learning).

No programming will be necessary for this course. Instead, the goal is to bring linguists from a variety of backgrounds in conversation with recent developments in computational modeling (and the excitement around their ‘abilities’). Supplemental interactive code notebooks may be circulated for those interested in engaging more deeply with the computational experiments highlighted in the course.

For those enrolled there are three requirements, i) active participation in class discussions, ii) posting comments/questions/thoughts on readings via canvas, and iii) a squib to be submitted at the end of the course. Visitors are welcome – either regularly or sporadically! Please send me your email address if you are not registered so that I can add you to the canvas.

Quoting from Giorgio’s quote from Amir and Kai’s quote of Bob’s seminar description: “The schedule will be flexible and open ended, following the discussion where it leads, and spending as much time on each topic as it seems to need. That is, we will make it up as we go along.”

24.S95: Special topics: Linguistics in K-12 education

• Instructor: Maya Honda
• Wednesdays, 2-5pm
• Room: 26-142

In this seminar, we will explore the idea that the study of language in K-12 (kindergarten-grade 12) education can be a means to develop young people’s understanding of scientific inquiry as well as their understanding of the nature of language. We will examine the view that the native language knowledge that each student brings to the classroom comprises a rich, accessible database, which can be used to give students the opportunity to become familiar with the methods, concepts, and attitudes of scientific inquiry. We will probe past and current efforts to engage young learners in linguistic inquiry and consider how to advance this work.

The challenge of this seminar is to create pedagogical materials and methods that will motivate learners of all ages to be inquisitive about their native language and about language in general, with a primary focus on secondary school students (grades 6-12). Seminar participants will work with one another and in partnership with K-12 teachers whenever possible.

There are two prerequisites for the seminar: the first is that you come motivated to making linguistic inquiry accessible to all and the second is that you come committed to collaborating with others in this work. Previous experience teaching linguistics at any level is welcome, but not required. Graduate students from other departments and undergraduates are also welcome if they have taken a linguistics course or if they have the instructor’s approval.

We are delighted to announce the publication in Glossa of (dissertating student) Peter Grishin’s paper “Omnivorous third person agreement in Algonquian”.  Congratulations Peter!! Here’s the abstract:

I argue that third person is not underspecified: there must be a distinct third person feature. I add to the existing body of morphological arguments for this conclusion (Nevins 2007; Trommer 2008, a.o.) a syntactic argument: I show that there is omnivorous third person agreement in Algonquian languages. I focus here on two, Blackfoot (Plains Algonquian) and Plains Cree (Central Algonquian), demonstrating that they have an agreement suffix (the peripheral suffix, analyzed as a probe in C) that indexes the number, animacy, and obviation of the structurally-highest third person argument, skipping over first and second person if it has to. I argue that alternative analyses of this agreement pattern in terms of animacy, obviation, and the categorial feature [D] do not work; thus, third person must be specified even in the syntax (contra Preminger 2019).

Glossa is an open-access journal so everyone can access the paper here: https://www.glossa-journal.org/article/id/8874/

Speaker: Ksenia Ershova (MIT)
Title: Manifestations of syntactic ergativity: disassociating high absolutive syntax from the ergative extraction constraint
Time: Tuesday, December 13, 1pm - 2pm
Location: 32-D462

Abstract: Syntactic ergativity is broadly defined as the sensitivity of syntactic rules to the distinction between subjects of transitive verbs (= ergative) on the one hand and objects of transitive verbs and subjects of intransitive verbs (= absolutive) on the other hand. The defining property of syntactically ergative languages is taken to be the ergative extraction constraint: while absolutive arguments are accessible for extraction, A’-movement of the ergative argument is ungrammatical (Deal 2016; Polinsky 2017, a.o.). A prominent analysis of syntactic ergativity involves the movement of the absolutive object to a position above the ergative agent (Aldridge 2004,2008; Coon et al. 2014,2021; Tollan and Clemens 2021, a.o.). In this talk, I argue that high absolutive syntax does not straightforwardly predict the ergative extraction constraint. Despite appearing to be a drawback of this approach, I claim that this is a desirable aspect of the theory. I support this claim with data from West Circassian, a language that displays syntactic ergativity in a range of grammatical domains but does not display a ban on ergative extraction.

Prosody and Identity in Linguistic Variation: Minicourse

Times: 12/6 and 12/8, 12:30pm-2pm
Location: 32-D461

Variation in intonation and voice quality is among the least well-described phenomena in both phonetic and sociolinguistic research. This is especially a limitation due to research in recent years that has described the importance of such variables for speakers and listeners in presenting and interpreting social-indexical information (Purnell et al 1999, Thomas and Reaser 2004, Holliday 2021). This body of research has also shown that intonation and voice quality variables can carry a range of social meanings that can be controlled and manipulated by speakers both at and below the level of consciousness. Thus, we begin with the premise that speakers’ use of pitch, prosody, and voice quality variables is a robust area to investigate the intersection of language and identity, as well as the nature of phonetic variation. In this minicourse, participants will be introduced to basic concepts in sociophonetics, with special attention to prosodic and voice quality variables. The course will concentrate on the theory and application of methods that help to elucidate how speakers communicate various aspects of personal identity, as well as how listeners perceive them. These fundamental aspects range from the community-level, such as region and race/ethnicity, to the more individual-level, such as persona construction.

On the first day of the minicourse, learners will be introduced to the major questions of sociolinguistics in the 21^st century, as well as sociophonetic methods and analytic procedures, with a focus on prosodic and voice quality variables. The readings for this day come from Erik Thomas’ influential textbook, Sociophonetics: An Introduction (2011), in order to provide students will an equitable starting point for better understanding such analyses. On Day 2 of the minicourse, participants will read two journal articles that focus on different types of questions in sociophonetics. The first paper, Thomas and Reaser (2004) is a foundational work that describes several decades worth of research on the issue of ethnic identification in sociolinguistics. The second paper, Burdin, Holliday, and Reed (2022), is a more recent study focusing on describing production-level differences between speakers of three different lects of American English. Participants will leave with a better understanding of how prosody works in sociolinguistic variation, which will enhance their ability to ask questions about variation that they may encounter in their own research.

References:

Burdin, R.S., Holliday, N.R. and Reed, P.E., 2022. “American English pitch accents in variation: Pushing the boundaries of mainstream American English-ToBI conventions”. Journal of Phonetics, 94, p.101-163.

Holliday, N.R., 2021. “Perception in black and white: Effects of intonational variables and filtering conditions on sociolinguistic judgments with implications for ASR”. Frontiers in Artificial Intelligence, pp.102.

Purnell, T., Idsardi, W. Baugh, J. 1999. “Perceptual and Phonetic Experiments on American English Dialect Identification.” Journal of Language and Social Psychology 18 (1): 10–30.

Thomas, E.R. and Reaser, J., 2004. “Delimiting perceptual cues used for the ethnic labeling of African American and European American voices”. Journal of sociolinguistics, 8(1), pp.54-87.

Thomas, E.R. 2011. Sociophonetics: An introduction. Basingstoke and New York: Palgrave Macmillan.

Title: Sociophonetic Variation and Human Interaction with Digital Voice Assistants

Location: 32-141
Time: 3:30pm-5pm, 12/9

Abstract: As technology that relies on speech is increasingly integrated into modern American society, voice assistants are becoming a more significant part of our everyday lives. This talk will present the results of three studies that focus on social perception of voice assistants, voice quality variation among the assistants themselves, and how one assistant’s “tone of voice” evaluation reinforces systematic linguistic bias. Results of the first study demonstrate how listeners engage in racialized judgments of digital voice assistants and that these judgments interact with perceptions of the personality of such assistants, providing evidence that listeners personify these voices. Results of the second study shed light on the voice quality features that may trigger judgments of speaker race and personal characteristics, even when the speaker is non-human. Finally, results of the third study show the ways in which speech recognition technology can reinforce and perpetuate bias against already marginalized groups of speakers. A more comprehensive understanding of how sociolinguistic variation interacts with the design of such assistants may help us to understand how listeners process variation and make judgments of voices, both digital and human. Additionally, a thorough analysis of how computational systems police speaker behavior can help us address systematic inequality as the linguistic line between humans and computers becomes increasingly porous.

Speaker: Omri Doron and Jad Wehbe (MIT)
Title: Post-Accommodation Informativity
Time: Wednesday, December 7th, 1pm - 2pm

Abstract: We discuss a constraint on global accommodation proposed by Heim stating that an accommodated presupposition must not settle the QUD (Heim, 2015). We argue that this constraint follows from the asymmetry between the pragmatic status of presuppositions and assertions assumed by the satisfaction theory (Stalnaker, 1970; Kartunen, 1974). We provide evidence that local accommodation is sensitive to this constraint, thus arguing that local accommodation maintains the asymmetry. This poses a challenge to theories of local accommodation.

Speaker: Aljoša Milenković (Harvard University)
Title: Markedness is not enough: Tone-stress interaction in Optimality Theory revisited
Time: Monday, December 5th, 5pm - 6:30pm

Abstract: As a general tendency in many languages, higher tone attracts/is attracted by metrical prominence, while lower tone tends towards metrically weak positions (Goldsmith 1987; Hayes 1995; Smith 2002; de Lacy 2002). The existing Optimality-Theoretic accounts use either negative markedness constraints or prosodic licensing to model tone-to-foot mapping. In this talk, I argue that both markedness- and licensing-based approaches fail to capture the full range of cross-linguistic variation. The markedness-based approach (de Lacy 1999, 2002, 2007) faces two empirical problems. First, it predicts a universal dispreference for higher tone in the weak position of a foot. This prediction is at odds with the stress pattern of Neoštokavian (Standard) Serbian, a South Slavic dialect with tone-driven stress, which preferentially constructs disyllabic trochees with a High-toned nonhead (Bethin 1994, 1998; Zsiga & Zec 2013). Second, given that negative markedness constraints penalize Low and Mid tone in stressed syllables, the theory treats stressed contour tones as marked because of the markedness penalty incurred by their Low/Mid-toned components. Consequently, contour tones are expected to be avoided and/or eliminated under stress. This is inconsistent with the fact that many languages restrict contour tones to stressed syllables, and no language restricts contour tones to unstressed syllables (Zhang 2000, 2001). The licensing-based account (Breteler 2017, 2018; Breteler & Kager 2022) readily explains both the preference for High-toned foot nonheads observed in Serbian and the metrical behavior of contour tones. However, unlike negative markedness constraints, the licensing approach has no means to enforce higher tone in foot heads and lower tone in foot nonheads, thus missing a well-documented empirical generalization. As a solution, I pursue a hybrid approach which combines de Lacy (2002)’s *Nonhead-Tone hierarchy with the licensing constraints of Breteler (2018). The midway approach adopted here is shown to improve the typological coverage of both existing approaches.

who: Dr. Sherry Yong Chen
when: 11/30, 2pm
where: virtual talk (contact Hadas for zoom link)
what: Sherry got her PhD in linguistics from MIT, where she worked on semantics/pragmatics and language acquisition. She currently works as a Knowledge Engineer at Amazon - Alexa AI. In her role, she creates ontological designs to expand Alexa’s question-answering abilities, performs data analyses to understand existing gaps, and does program management for her team’s event coverage. Most recently, she worked on Alexa’s QnA coverage for the midterm election.

To learn more about her team: https://www.amazon.jobs/en/locations/sba-california

Sherry came to the US as an international student, and has experience in navigating the job market and the immigration processes.

Speaker: Norvin Richards (MIT)

Title: Finding Something to Lean On

Date & Time: Thursday 12/1, 12:30—:40PM

Location: 32-D461, https://mit.zoom.us/j/96057548137

Abstract: 

Our local homogeneity expert Jad Wehbe (MIT) gave a very well-received talk on 11/19 at the online workshop on Homogeneity and Non-Maximality in Plural Predication and Beyond!

His talk was titled Revisiting presuppositional accounts of homogeneity: 

Later in the day, he also served as a panelist on the panel Homogeneity projection and relation to other gappy phenomena: 

We’re excited to welcome Jad this week! The meeting will be in person, back to our original venue, and we’ll set up OWL for people in the department who want to attend virtually.

Zoom link: https://mit.zoom.us/j/94298987190
Speaker: Jad Wehbe (MIT)
Date and time: Wednesday 11/16, 1-2pm
In-person location: 32-D461

Title: Revisiting presuppositional accounts of homogeneity
​
Abstract: Early accounts of homogeneity effects with definite plurals treated homogeneity as a presupposition (Schwarzschild, 1996; Löbner, 2000; Gajewski, 2005), but this characterization has recently been challenged on the basis that homogeneity does not seem to exhibit the standard projection patterns commonly attributed to presuppositions (Spector, 2013; Križ, 2015). The goal of this talk is to argue that homogeneity is in fact a presupposition, despite the apparent differences. In the first part of the talk, I will show that homogeneity is sensitive to a constraint on presupposition accommodation proposed by Heim (2015). Taking this constraint as a diagnostic of presuppositionality, this provides empirical evidence that homogeneity is a presupposition. In the second part of the talk, I will argue that the only difference between homogeneity and standard presuppositions are the conditions under which they can be locally accommodated. I discuss a related view proposed by Fox (2017) and argue that the differences in projection follow from the fact that homogeneity gives rise to non-connected propositions (in the sense of Engeuhard and Chemla (2021), while standard presuppositions are generally connected. ​

Norvin Richards and Roger Paul are featured on CBC on their efforts to revitalize the Wolastoqey (what used to be called Maliseet) language! 

https://www.cbc.ca/news/canada/new-brunswick/norvin-richards-roger-paul-mit-wolastoqey-linguistics-language-new-brunswick-1.6637660​

who: Harry Calvert (Speechmatics)
when: Friday, 11/11, 9am (note unusual time/date)
where: virtual talk (contact Hadas for link), there is no in-person room for this event
what: We will meet with a recruiter from the startup Speechmatics. Speechmatics is interested in establishing a relationship with MIT linguists and has interesting opportunities for internships as well as full-time employment. We will use this opportunity to ask Harry about the role of a recruiter in general, to get a recruiter’s perspective on the hiring process, including tips and common mistakes they see. Here is a short blurb from Harry:

Speechmatics is a deep learning start-up that has developed the most accurate and inclusive Speech-to-Text engine in the World. Recently raising $62mil in series B funding, Speechmatics is on a mission to ‘understand every voice, in every context’.

Harry Calvert is a Global Talent Partner at Speechmatics with over 7 years experience in Recruitment; attracting some of the smartest minds in AI to conduct cutting-edge research, whilst building relationships with external partners/universities.

who: Dr. Khia Johnson
when: Wednesday, 11/9, 2pm
where: virtual talk (contact Hadas for zoom link), or 5-231
what: Khia Johnson has a PhD in linguistics from the university of British Columbia and currently works as a UX Research Scientist on the Reality Labs Research Audio team at Meta. Her research in graduate school focused on corpus and experimental phonetics and psycholinguistics. In her current role, she conducts behavioral and attitudinal human subjects research to evaluate novel audio technology from the perspective of humans.

For more information about her team’s work, see:
https://about.fb.com/news/2020/09/facebook-reality-labs-research-future-of-audio/
https://www.cnet.com/tech/computing/behind-the-doors-of-metas-top-secret-reality-labs/
https://www.khiajohnson.com/

Date and time: Wednesday 11/02, 1-2pm
In-person location: 32-D461
Zoom link: https://mit.zoom.us/j/94298987190

Title:  ”Quasi-ECM” constructions in Modern Greek: Evidence for semantic lowering

Abstract:

MG displays certain attitudinal constructions where an attitude verb may take an accusative object (henceforth ACC DP) followed by a subjunctive CP. Kotzoglou and Papangeli (2007) dub this the “quasi-ECM” construction. The basic pattern is the following:

(1)
I Maria theli ton Yani /o Yanis [na tin aghapai].
The.nom Maria.nom want.prs the.acc Yani.acc /the.nom Yani.nom [comp her.acc love.subj].
`Maria wants John to only love her.’

Hadjivassiliou et al. (2000); Kotzoglou and Papangeli (2007); Kotzoglou (2013, 2017) provide considerable evidence that the ACC DP is base-generated in the matrix clause. We provide additional arguments that this is indeed the case, and we argue that these are proleptic constructions as opposed to object control​. Contrary to what is usually assumed in the literature, given the right context, the pro can be in object position too:

(2) Context: Yanis is a political activist and part of an organization run by me. I want to raise awareness about
the organization and I think that getting someone arrested will give us some publicity to this end. 

Thelo ton Yani [na ton silavi i astinomia].
Want.prs the.acc Yani.acc [comp him.acc arrest.subj the.nom police.nom].
`I want the police to arrest Yanis.’

Cross-linguistically, prolepsis marks a de re (e.g. German (Salzmann, 2017), Nez Perce (Deal, 2018)) or a de re and a third reading (e.g. Tiwa (Dawson and Deal, 2019)). We argue that in MG de dicto readings are also possible:

(3) Context: Little Petros is in kindergarten and he and his friends believe that green dogs exist. One day they are talking about green dogs and Petros bets that exactly three of them will show up at his party.

O Petrakis theli akrivos tris prasinus skilus [na erthun sto parti].
The Petros.dim want.prs exactly three.pl green.acc.pl dog.acc.pl [comp come.subj.pl in-the party]. 
`Little Petros wants exactly three green dogs to come to the party.’

This attitude report does not commit the speaker to the existence of green dogs; in Fodor’s terms (Fodor, 1970), the embedded subject is read opaquely. Thus, proleptic constructions are not always interpreted transparently, contrary to what has been assumed up to now. 

We provide an analysis in terms of semantic lowering, as well as argue that prolepsis in Modern Greek still has more restricted truth conditions than its non-proleptic equivalent derived from movement. Namely, there is a requirement that the DP is part of what causes the CP to happen. So, (1) with the ACC is only felicitous if Yanis takes some kind of action to m​ake the CP hold. ​This suggests that prolepsis is not simply a mechanism to exclude de dicto readings, but a way to express some marked meaning in general (which we formalise for Modern Greek). Finally, we hint at an implicational hierarchy of prolepsis, suggesting that if a language has a de dicto reading, then it also has a third and a de re one.

Speaker: Ido Benbaji & David Pesetsky (MIT)
Title: E-Extension and the Uniformity of Silence
Time: Thursday, November 10th, 12:30pm - 1:50pm

Abstract: At the heart of this talk is the conjecture that all syntactic processes that silence an otherwise overt expression do the same thing: apply an E-feature to that expression (Uniformity of Silence):

(1)
Effect of E-feature
The feature E associated with an occurrence of node α silences every element reflexively dominated by α.

With this conjecture as background, we focus here on the silencing of syntactic heads and argue that at least some phenomena traditionally studied under the special rubric of ellipsis involve nothing more than the extension of an E-feature applied for independent reasons to a slightly higher node:

(2)
E-Extension
Optional: If Hº bears E, copy E to the smallest phrasal node containing H and its selected complement.

We propose that the phenomenon known as Sluicing is simply the result of the E-feature applied by an independently motivated rule (optional in some languages) to the complementizer in a “Doubly-Filled Comp” configuration, optionally extended to the minimal C′ by (2).

Furthermore, the family of phenomena usually studied under the rubric of VP-Ellipsis, including English Auxiliary Verb Phrase Ellipsis, are simple the result of the E-feature that silences the trace of verb raising optionally extended to the minimal V′ or AUX′ containing that verb by (2). This dovetails with proposals by Thoms (2010) and Harwood (2014), among others, which also posit verb movement as a necessary precursor to “VP ellipsis” (in somewhat different ways)

Both phenomena crucially target a phrase minimally larger than the silenced head, yielding the phenomenon dubbed Adjunct Exclusion in recent work (including Landau, passim). We argue that “Adjunct Exclusion” is a subcase of a more general phenomenon of non-complement exclusion observed in both of the ellipsis phenomena studied here, with evidence from English and Hebrew.

Speaker: Katie Martin (MIT)
Title: A new presuppositional account for slurs
Time: Wednesday, November 2nd, 1pm - 2pm

Abstract: Preexisting analyses of slurs have treated the negative appraisal associated with their meaning as deriving either semantically, via presupposition or conventional implicature (Cepollaro (2015) and Gutzmann (2015), among others), or pragmatically, via Gricean reasoning (Nunberg (2018) and Bolinger (2017) among others). I respond in this project to two recent works that have criticized these analyses of slurs, both semantic (Lo Guercio 2021) and pragmatic (Falbo 2021).

Falbo (2021) observes that accounts that deem slurs to be semantically equivalent to their neutral counterparts are significantly challenged by what she calls “non-target” uses, such as the following:
(1) Deandra’s a lesbian, but she’s not a dy*e.
(2) Not only is my neighbour a lesbian, but she’s a total dy*e too. She drives a motorcycle and is covered in tattoos.

Under the claim that slurs and their neutral counterparts have identical semantics, (1) should be incoherent and (2) should be redundant – however, this is obviously not the case. Indeed, this data provides a challenge not just for pragmatic accounts, but for semantic accounts as well – if the only difference in meaning between “lesbian” and “dyke” is that the latter presupposes a negative attitude towards lesbians on the part of the speaker, it seems distinctly odd that a single speaker would utter both of the clauses in each of (1) and (2) – and yet such utterances are perfectly natural.

In a similar vein, Lo Guercio (2021) criticizes presuppositional accounts of slurs on the grounds that even bigoted speakers can use “neutral” forms in place of slurs, which ought to cause a violation of Maximize Presupposition, but seems in fact to be perfectly felicitous.

I address these important objections by proposing a theory of slurs in which the presuppositional content associated with the use of a slur is not simply a negative attitude on the speaker’s part towards the group identified by the slur, but rather a negative attitude towards the stereotypes associated with that particular group and, crucially, that these stereotypes are relevant/salient in the conversational context.
(see also abstract attached)

Speaker: Brian Leahy (Harvard)
Title: Problems with Possibilities
Time: Thursday, November 3rd, 12:30pm - 1:50pm

Abstract: I will present the results of several studies that test preschoolers’ ability to talk about and think about mutually incompatible possibilities. These studies reveal systematic errors in how preschoolers (a) answer questions about what can and has to happen, and (b) form contingency plans when choosing actions in the face of multiple open possibilities. For example, when choosing between a container that must contain a prize and two additional containers that merely might contain a prize, older 2-year-olds choose a risky option half of the time. They take a risk when they could have a sure thing. This error slowly becomes less frequent with age, but is still common among 4-year-olds.

In this talk I will quickly sketch some of the behavioral results, and put an explanation on the table that accounts for the observed patterns of errors: when faced with multiple incompatible possibilities, children use simulation to generate one possible outcome and then treat that outcome as the fact of the matter. They do not mark the simulated outcome as merely possible and check whether there are other possibilities that need to be accounted for.

After sketching the behavioral results, I will present our language comprehension studies. There is no doubt that learning the English modal auxiliary system is difficult: each modal auxiliary can express many different meanings, and many different auxiliaries can express the same meaning. I will present evidence that comprehension of modal vocabulary appears late, around the same time that children start succeeding on the behavioral tasks described earlier. I will argue that the complexity of the mapping problem cannot fully account for children’s struggles. In particular, I will show that there are correlations in children’s ability to answer questions about what has to happen and their ability to apply nonmodal vocabulary like ‘all’ and ‘only’ in the modal domain. Children comprehend ‘all’ and ‘only’ by age 4, so these correlations are surprising if children’s struggles with modal auxiliaries arise only from the complexity of the mapping problem.

What is the additional problem that children face? Why is there such consistency across age in children’s ability to talk about and think about mutually incompatible possibilities? One candidate answer is that the ability to mark representations as merely possible emerges late in the preschool years, typically after age 4. Existing data do not rule out this hypothesis, and it offers a tidy explanation for all of the observed data. This motivates a direction for future research: If the ability to mark representations as merely possible is one that emerges with age, what are the mechanisms by which it emerges?

Who: Sarah Clark
when: 2pm
where: virtual talk (contact Hadas for zoom link)
what: I completed my masters in linguistics in 2019 at the University of Illinois at Urbana-Champaign. At the moment, I am working toward my doctorate in linguistics at UIUC. I specialize in sociolinguistics, specifically critical discourse analysis, the interplay of ideology and identity, organizational communication, and military and veteran identity. I have been working during my whole program for the US Army Corps of Engineers’ Construction Engineering Research Laboratory. My duty title is Strategic Communications Officer and I lead all internal organizational communication to support and grow our multi-million dollar portfolio. In other words, I help engineers talk to each other and non-engineers in order to do the mission. Working outside of academia is by far the gold standard for linguists in my opinion - we are linguists who are creative, resilient and have the honor of bringing the power of linguistics to the real world!

MIT folks presented at the 2022 Annual Meeting on Phonology (AMP 2022) which happened on October 21-23 at UCLA. Donca gave an invited plenary talk, and students and faculty members gave presentations: 

Current students and faculty:

Donca Steriade (Faculty): Vowel-to-Vowel intervals in quantitative meter (Plenary Talk)

Boer Fu (6th year): The Segment Status of the Mandarin Glide: A Language Game Experiment (poster)

Eunsun Jou (4th year): An economy-based amendment to Robust Interpretive Parsing with the GLA (poster)

Giorgio Magri (Visiting Associate Professor, PhD 2009) and Arto Anttila: Paradoxes of MaxEnt markedness (poster)

Canaan Breiss (Postdoctoral Associate, MIT BCS): Lexical Conservatism as an empirical challenge to obligatory cyclic inheritance (talk)

Alums: 

Ora Matushansky (PhD 2002): Russian transitive softening as ablaut

Bingzi Yu and Youngah Do (PhD 2013): The transmission of vowel (dis)harmony: An iterative learning study

Coppe Van Urk (PhD 2015) and Adam Chong: The roots of non-concatenative morphology in Western Nilotic

Some of MIT presented at the 54th Algonquian Conference this weekend (Oct 20-23) at CU Boulder! 

Peter Grishin (5th year): Subordinative long distance agreement in Passamaquoddy-Wolastoqey and the syntax of the inverse

Peter Grishin (5th year) and Will Oxford (visiting associate professor): Patterns of portmanteau robustness across Algonquian

Giovanni Roversi presented at NYU’s Ling-Lunch on Oct 13, 2022! 

Title: Where can probes be? Evidence from asymmetric adjectival concord

Abstract: We currently don’t have a theory telling us what kind of probes should or shouldn’t be where; I’m not going to propose one either. In this talk I will make an indirect argument for not wanting too restrictive a theory: the empirical landscape, when looked at carefully, is too varied for us to be able to afford a restrictive theory of probe distribution. The domain I will concentrate on is asymmetric adjectival concord, that is, languages where the morphological patterns on attributive adjectives and predicative ones (“the red car” vs “the car is red”) are different. Concretely, I will look at German, North Sámi and Northern Norwegian, which I argue exhaust the logically possible space of variation for the matter in question (that is: (i) only concord on attributive but not on predicative; (ii) the opposite of that; (iii) concord on both but of different kind). I will try to convince you that especially the North Sámi and Northern Norwegian patterns are real, and will exclude alternative analyses. In other words: the typology of attested patterns does in fact cover all logically possible asymmetries, and our theory should therefore be able to derive all of them.

On August 12, 2022, we had a beginning of term party @ Mex, seeing old and new faces with smiles all around! 

Speaker: Donca Steriade (MIT)
Title: Vowel-to-Vowel intervals in quantitative meter
Time: Monday, October 17th, 5pm - 6:30pm
Location: 32-D831

Abstract: 

I present evidence for V-to-V intervals as units of rhythm, drawn from laws that govern the distribution of word-final V̆C0# sequences in the quantitative meters of the classical languages.

Weight-changes in phrasal contexts: The talk analyzes changes in the weight of word-final V̆(C)# positions between phrase-final and phrase-medial contexts, as diagnosed by the quantitative meters of Greek and Latin. The relevant generalizations are outlined in (2). The top row shows the metrical weight of the first position in a V̆CC0V string containing a word boundary, as observed in dactylic hexameter lines. The 3rd and 4th rows indicate that some phrase medial cross-boundary V̆CC0V sequences are underattested, thus arguably disfavored.       

The restricted shaded cells contain sequences that are underattested at major junctures in the dactylic hexameter, or absent in all positions. Unshaded cells correspond to favored sequences (V̆C#V is favored over V̆#CV at a major juncture) or unrestricted ones (V̆CC#V compared to V̆CC#CV). Unlike V̆C# and V̆# finals, heavier V̄C0 # and V̆CC(C)# finals are not restricted.

Interval-based and syllable-based analyses: The sequences in (2) are parsed into intervals in table (3), and into syllables in table (4). Shaded cells correspond, as in (2), to restricted sequences:

The interval parses in (3) reveal a simple reason for selective underattestation: in all shaded cells, the weight of the word-final interval has changed in the shaded cells relative to its weight in isolation, as one or more Cs were added to it by the following word. By contrast, the word-final interval in (a) and (e) maintains line-internally the same weight as in isolation.

On an interval analysis, all patterns of underattestation outlined in (2-4) emerge as driven by the preference for weight correspondence between intervals in isolation and their correspondents in line-medial or phrase-medial contexts. All substantial changes of weight between the isolation context and the line-medial context are disfavored, and cause poets to avoid the sequence.

Compare now the syllable-based parses, (4). No syllabic factor distinguishes restricted sequences from unrestricted ones. The weight of the first syllable in (4.a) and (4.c) must be attributed, on any syllable-based theory of weight, to the reassignment of a C from one syllable to a neighboring one, turning heavy final V̆C.# into the initial light V̆.C#V in (4.a); and light V̆.# into heavy V̆#C. in V̆#C.CV, (4.e). So both (4.a) and (4.c) have undergone resyllabification, but only (4.c) is restricted. Also hard to explain is why, of the two sequences that begin with a light syllable, VC#V (4.a) and V#CV (4.b), the unrestricted sequence V.C#V is the result of resyllabification. Similarly mysterious is why (4.d) V̆C.#CV, with syllable and word boundaries properly aligned, is restricted, while syllabically misaligned (4.e) V̆C.C#V occurs freely.

Speaker: Paul Rössler (Universität Regensburg)

Time: Thursday 10/20, 12:30-1:50pm

Zoom link: https://mit.zoom.us/j/98155777682

In-person location: 32-D461

Title: Comma and prefield. What does German orthography tell us about the relation between norm, system and language use?

Abstract:

In text types with distinct nominal style, such as academic texts, the writer tends to mark the extensive complex nominal phrases in front of the finite verb graphically by comma placement. Meanwhile, this phenomenon seems to have outreached the domain of academic texts. The talk reveals how widely this writing practice which is incompatible with the norms of standard German orthography has already spread. This raises the question how linguists and codifiers deal with that pervasive phenomenon of language standardization as players in the field of social forces, and which consequences this entails for teachers as norm authorities and norm mediators.

Professor Rössler is co-PI with Kai von Fintel on an MISTI seed grant on “Comparative Punctuation” and visiting MIT for a couple of weeks.

The “careers for linguists” workshop will host our very own Ruth Brillman (PhD 2017), who is currently an engineering manager for Google Search and works right around the corner from us. She’ll join us in person for a Q&A about her path to industry.

When: Wednesday, 10/12, 2-3:30pm
Where: 32-D461 or on zoom (email me for a link)

“I graduated from MIT with a PhD in Linguistics in 2017, where I specialized in syntax (primarily) and language acquisition. I’ve spent the past five years in industry, mostly working on NLP projects and building software informed by a structural understanding of language. Currently, I’m a software engineer at Google where I lead a small team focused on improving Google Search quality in lower-resource languages.”

Speakers: Meg Gotowski (MIT) & Kristen Syrett (Rutgers)

Time: Thursday 10/13, 12:30-1:50pm

Zoom link: https://mit.zoom.us/j/96548700087

In-person location: 32-D461

Title: It is daxy to learn adjectives, and learning adjectives is daxy for everyone: Syntactic frames support the acquisition of adjective meaning.

Abstract: How do children learn the meaning of adjectives like fun or delicious? These are adjectives that are context-dependent and speaker-dependent, encoding both gradability and subjectivity. For this reason, subjective such as these adjectives pose a challenge for word learning, given that they do not have a fixed referent. An influential theory in the word learning literature is syntactic bootstrapping, which claims that learners recruit the syntactic environment in which a word is found in order to deduce its meaning (Landau & Gleitman 1985). While syntactic bootstrapping has been examined extensively in regard to verb learning (Gleitman 1990; Fisher 2002; Gleitman et al. 2005), adjectives have received considerably less attention—and previous studies have also conflated syntactic and semantic information (such as animacy, Becker et al. 2012). In this research, we focus on subjective adjectives as a case study for analyzing the influence of the syntax alone within the adjectival domain. We discuss the results of a word learning experiment modeled after the Human Simulation Paradigm (Gillette et al. 1999), during which participants were presented with a novel adjective in a set of syntactic frames. Each set of frames crucially reflected the unique distributional signature of five different subclasses of subjective adjectives (Bylinina 2014). Participants demonstrated a sensitivity to these frames, and reliably recruited the syntax in order to narrow down the potential meanings for the novel adjectives, offering responses consistent with the set of frames provided—while becoming increasingly confident in their responses after encountering the adjective across multiple frames.

We’re excited to welcome Keely at the first LFRG of the semester! The meeting will be in person, but we’ll set up OWL for people in the department who want to attend virtually.

Zoom link: https://mit.zoom.us/j/94298987190​
Speaker: Keely New (MIT)
Time: Wednesday 10/12, 1-2pm
In-person location: 32-D461

Title: Associative plurals and the plural pronoun construction

Abstract: In Burmese, a conjunction of associative plural DPs has an available reading where the named individuals in the conjunction internally satisfy the plural requirement. I call this the internal plural reading. The internal plural reading is reminiscent of what has been called the plural pronoun construction in Russian, Greek etc. In this talk, I present a post-suppositional semantics for Burmese associative plurals. In a nutshell, I propose that the named individuals in a conjunction of associative plurals can satisfy each other’s plural requirement if the associative plural has a meaning that is post-suppositional. I discuss the extent to which this post-suppositional analysis can be extended to the PPC, particularly in light of the empirical landscape regarding the availability of PPC and internal plural reading across and within languages.​

Welcome to visiting students Janek Guerrini, Elham Mehr, and Nur Lan!

Janek Guerrini: I am a PhD student in linguistics and philosophy at Institut Jean Nicod, at Ecole Normale Supérieure in Paris, working under the supervision of Benjamin Spector. I specialize in semantics and philosophy of language, and my main research interests are non-intersective modification and genericity. Here’s why I think they’re interesting: both of these phenomena have posed a long-standing challenge to semantic theory, and both seem to be particularly sensitive to the structure of our conceptual, non-linguistic representations. This makes for a domain where the study of language can both illuminate and be illuminated by broader questions in cognitive science. Most recently, I have been looking at some generic readings of indefinites in constructions we use to talk about similarity, as in “look like a lawyer”, which are puzzling because they do not seem to be reducible to an adverbial notion of genericity. 

Elham Mehr

Nur Lan: I’m a visiting PhD student from ENS Paris, co-advised by Emmanuel Chemla (ENS) and Roni Katzir (Tel Aviv University). I originally studied computer science and I like to build models of language learning.

Speaker: Giovanni Roversi (MIT)
Title: Where can probes be? Evidence from asymmetric adjectival concord
Time: Tuesday, October 4th, 1pm - 2pm

Abstract: We currently don’t have a theory telling us what kind of probes should or shouldn’t be where; I’m not going to propose one either. In this talk I will make an indirect argument for not wanting too restrictive a theory: the empirical landscape, when looked at carefully, is too varied for us to be able to afford a restrictive theory of probe distribution. The domain I will concentrate on is asymmetric adjectival concord, that is, languages where the morphological patterns on attributive adjectives and predicative ones (“the red car” vs “the car is red”) are different. I will try to convince you that the typology of attested patterns does in fact cover all logically possible asymmetries, so that our theory should be able to derive all of them.

Meg Gotowski and Forrest Davis will be giving presentations on their dissertation research. Join us on Friday (9/30) from 2-3:30 in the 8^th floor conference room (32-D831). 

It is DAXY to learn! Bootstrapping in the Adjectival Domain (Meg Gotowski) 

Abstract: An influential theory in word learning is known as syntactic bootstrapping (Landau & Gleitman 1985), which claims that children are able to map structure to meaning.  Most of the bootstrapping literature has focused on the ability of learners to rely on syntactic frames in order to deduce the meaning of verbs (see Gleitman et al. 2005). In this talk, I examine how syntactic bootstrapping extends to the adjectival domain, focusing on how learners are able to acquire different subclasses of subjective gradable predicates (e.g. fun, tasty, tough). I discuss the results of an experiment based on the Human Simulation Paradigm (Gillette et al. 1999), and argue that while learners are sensitive to individual adjectival frames, they are also dependent on seeing adjectives across multiple frames in order to effectively narrow down the hypothesis space of possible meanings (consistent with Mintz 2003 for verbs).  

Neural Models of Language and the Limits of Superficialism (Forrest Davis) 

Abstract: A typical approach to evaluating neural models of language for linguistic knowledge will find instances of overlap between humans and models. This overlap is claimed to be evidence that our linguistic theories can be simplified. I will instead argue for a different approach to evaluating such models. I advance the position that neural models are models of “superficialism”, the worldview which asserts that all meaningful linguistic (and more broadly psychological) distinctions can be made on the basis of observing ordinary behavior. By assuming this worldview, the role of data in determining a neural model’s behavior is centered. I then show via two case studies (ambiguous relative clause attachment and implicit causality) that mismatches between neural models and humans follow from general properties of data. I conclude by suggesting that, to the extent that these really are general properties of data, models will always be sensitive to incorrect generalizations. 

MIT had a huge turnout at Sinn und Bedeutung 27 in Prague, 14-16 September 2022! 

Current students:

+ Lorenzo Pinton (with Maria Aloni): Sluicing and Free Choice
+ Adèle Hénot-Mortier: A dynamic alternative-pruning account of asymmetries in Hurford disjunctions
+ Ido Benbaji: (with Omri) Adversative only is only only; (with Yash and Filipe) The Logic of Hindi Co-compounds
+ Enrico Flor: Questions in non-distributive belief ascriptions
+ Jad Wehbe: Against the lexical view of cumulative inferences and homogeneity
+ Omri Doron: (with Ido) Adversative only is only only
+ Anastasia Tsilia (poster): “Quasi-ECM” constructions in Modern Greek: Evidence for semantic lowering

Alums:

+ Itai Bassi: Strict readings of logophors and the LF of anaphoric dependencies (with others) 
+ Jonathan Bobaljik: (with Uli Sauerland) About ‘us’
+ Uli Sauerland (invited): An Algebra of Thought that Predicts Key Aspects of Language Structure; (with Jonathan Bobaljik) About ‘us’
+ Pritty Patel-Grosz (invited): The search for universal primate gestural meanings
+ Yasutada Sudo: Against simplification: free choice with anaphora
+ Sam Alxatib: Necessary Free Choice and its theoretical significance

Presenting but not in the picture:

+ İsa Kerem Bayırlı (alumnus): UM2: A Generalization over Determiner Denotations
+ Yash Sinha (with Ido and Filipe): The Logic of Hindi Co-compounds
+ Filipe Hisao Kobayashi (with Ido and Yash): The Logic of Hindi Co-compounds

Speaker: Peter Grishin (MIT)
Title: Passamaquoddy subordinative clauses are TPs
Time: Tuesday, September 27th, 1pm - 2pm

Abstract: Algonquian languages are known for their distinct inflectional paradigms that have different syntactic distributions, something that has received quite a bit of interest in the theoretical literature (Campana 1996, Brittain 2001, Richards 2004, Cook 2014, Bogomolets, Fenger, and Stegovec 2022, a.o.). However, one inflectional paradigm/clause type has not received much (if any) attention: the subordinative, an Eastern Algonquian innovation. In this talk I’ll present some fieldwork and corpus data on the subordinative in Passamaquoddy, proposing that the seemingly-unrelated syntactic contexts it appears in—clausal complements to certain verbs and modal particles, some clausal coordinations, and polite imperatives—can all be unified if we take subordinative clauses to be TP sized, lacking a CP layer. While I think the broad picture I sketch is compelling, there are some loose ends and problems that will emerge—I’m looking for help with figuring out how to deal with them.

The annual ballot of the Linguistic Society of America is now open and members have until November 5, 2022 to cast their votes. There are a series of proposed amendments to the LSA Constitution and Bylaws and the LSA’s website provides some comments from members pro and contra. The ballot also includes the slate of candidates for various positions in the Society:

• Marlyse Baptista (University of Michigan) for
  Vice-President/President Elect
• Shelome Gooden (University of Pittsburgh) for Language
  Co-Editor
• Four candidates for two at-large seats on the Executive Commitee:
  • Melissa Baese-Berk (University of Oregon)
  • Michel DeGraff (Massachusetts Institute of Technology)
  • Sali A. Tagliamonte (University of Toronto)
  • Michal Temkin Martinez (Boise State University)

Speaker: Janek Guerrini (Institut Jean Nicod, ENS)
Title: Genericity in similarity
Time: Thursday, September 29th, 12:30pm - 1:50pm

Abstract: In this talk, I offer an account of similarity constructions involving ‘like’, such as ‘be like’ and ‘look like’. I argue that these constructions have two key properties. (1) The first is that similarity predication amounts to predication of overlap of salient properties: I analyse ‘John is like Mary’ as ‘John shares relevant properties with Mary’. This is motivated by the fact that there seem to be grammatical devices that single out precisely what properties are relevant, e.g. ‘With respect to personality, she’s just like her father’. (2) The second key feature of similarity talk is, I argue, that it involves inherent generic quantification. This explains a range of data: first, it accounts for the reading of indefinites embedded in ‘like’ Prepositional Phrases: ‘John looks like a lawyer’ is almost equivalent to ‘John looks like a typical lawyer’. Second, it accounts for narrow-scope and almost conjunctive readings of disjunction in the scope of ‘like’: ‘Mary looks like a lawyer or a judge’ is almost equivalent (on its most accessible reading) to ‘Mary looks like a lawyer and Mary looks like a judge’.

Peter Grishin, Cora Lesure, and Norvin Richards spent the weekend of September 10-11 in Fredericton, New Brunswick, at the Wolastoqey Elder and Youth Gathering (“Nankomiptasu: Wolastoqewiyik Carry Language to Future Generations”).  We heard presentations related to language revival and maintenance, including discussions of several exciting projects that are under way; we also had a chance to spend time with the elders as they sat around a table telling stories in Passamaquoddy-Wolastoqey late into the night.  Many thanks to Roger Paul (MS ‘20) for telling us about this event!

Let us all give a big welcome to the incoming cohort, ling-22! 

Runqi Tan: I’m from China. My research interests started with modelling phonological form and structure with optimization models, and I’m fascinated by the general principles that shape the language systems. I enjoy spending time with friends, reading science, history, biography, watching cartoons and going to musicals and concerts.

Taieba Tawakoli

Ukhengching Marma

Congratulations to Frank Staniszewski, who successfully and excellently defended his dissertation on August 25, 2022, titled Modality and Time in Logical Context! 

The dissertation develops a theory of neg-raising that unifies the phenomenon with existing theories of free choice and negative polarity items. The empirical focus is on “until”-phrases and on the neg-raising predicates “want”, “should”, and “be supposed to”. Predictions of the formal account are then examined in a language acquisition experiment.

Many congratulations to Christopher Baron, who successfully and excellently defended his dissertation on August 15, 2022, titled The Logic of Subtractives, or, Barely anyone tried almost as hard as me!

The dissertation analyzes the elements “almost and “barely”, proposing a formal analysis in which they are subtractive modifiers of quantifiers that via exhaustification result in exceptive meanings. The resulting theory is then used to examine the compositional structure of comparative and equative constructions as well as numeral constructions.