Whamit!

Speaker: Shumian Ye (MIT/Peking University)
Title: Negated disjunctions in Mandarin revisited
Time: Wednesday, May 1st, 1-2PM
Location: 32-D461

Abstract: In this talk, I will argue that OR in Mandarin is neither an NPI nor a PPI, contra Crain (2012). First, when embedded by negations, certain disjunctions obligatorily take the narrow scope, while Hurford disjunctions obligatorily take the wide scope. Compare (1) with (2).

(1) ta mei quguo Beijing huo qita chengshi.
she NEG have.been Beijing OR other cities
Unique reading: She hasn’t been to Beijing or any other cities. (¬ > ∨)

(2) zhe ping jiu de jiage bu chaoguo shi kuai huo ershi kuai.
this CL wine DE price NEG exceed ten dollars OR twenty dollars
Unique reading: The price of this wine doesn’t exceed $10, or doesn’t exceed $20. (∨ > ¬)

Second, most negated disjunctions are ambiguous in Mandarin. Which reading is salient seems to depend on the higher modals and the contexts. Compare (3) with (4).

(3) ta kending mei quguo Beijing huo Shanghai.
she definitely NEG have.been Beijing OR Shanghai
Salient reading: She definitely hasn’t been to Beijing, nor to Shanghai. (¬ > ∨)

(4) ta keneng mei quguo Beijing huo Shanghai.
she maybe NEG have.been Beijing OR Shanghai
Salient reading: She maybe hasn’t been to Beijing, or hasn’t been to Shanghai. (∨ > ¬)

More crucially, the disjunctions can still take wide scope when the negated disjunctions are in the scope of a downward entailing operator, that is, OR in Mandarin cannot be rescued as a PPI (Szabolcsi 2004). After going through these arguments, I will take other modals into consideration and give a preliminary explanation for the contrast between (3) and (4), which mainly relies on two factors: 1) Interaction between different modals and ignorance/uncertainty inferences, 2) Competition between OR and AND.

Speaker: Michael Kenstowicz (MIT)
Title: Phonetic Correlates of the Stop Voicing Contrast in Javanese
Time: Wednesday (5/1), 5:00pm-6:30pm
Location: 32-D831

Abstract:
Previous studies of the Javanese [±voice] distinction find that in word-initial position the contrast is realized not in virtue of stop closure voicing or VOT but rather in the quality and pitch of the following vowel. Correlates include differences in the first and second formants as well as fundamental frequency and voice quality. In this study we extend this line of research by looking at the realization of the voicing contrast in a wider variety of contexts (word-initial, medial, and final) and tests of statistical significance of the data. The implications of our findings for the phonology of the language are also considered.

Speaker: Kyongjoon Kwon (Sungkyunkwan University)
Title: A phonologically null verb give in the Russian threat dative construction
Time: Thursday, 5/2, 12:30-1:50pm
Location: 32-D461

Abstract:
The central goal of this talk is to argue that the so-called Russian threat dative construction contains a phonologically null verb give. In the proposed ditransitive structure, the theme is an event denoted by a pronominalized verb that anaphorically refers to the verb phrase in the immediately previous speech. The pronominalized verb stem is phonologically realized through the incorporation of the null verb after a series of head movements. Under this proposal, the seemingly unmotivated or stipulated at the least dative case marked pronoun is well accounted for, i.e., as a recipient in the predication denoted by the ApplP. And the sentential meaning “I will give you/him/her/them back X ”, where X repeats a part of the previous speech, gives rise by implication to the threat effect. Finally, I propose that the posited null verb give, alongside overt forms, should be termed as “retortative” give, which expresses the idea of paying the interlocutor back with a similar response by retaliation.

Speaker: Keny Chatain (MIT)
Title: What is wrong with doubles?
Time: Wednesday, April 24th, 1-2PM
Location: 32-D461

Abstract: In this talk, I show that for a sizeable class of adjectival operators (including comparatives, superlatives, excessives, same, different), standard and otherwise perfectly adequate denotations and LFs deliver problematic truth-conditions in sentences where two such operators occur, e.g. (1-2). This happens even though speakers recognize such sentences as well-formed and have consistent intuitions about their truth-conditions. These problems have been partially tackled in the literature (von Stechow (1984) under the label multihead comparatives, Meier (2000) for degree result clauses, and refs therein), but a one-size-fits-all solution, if it exists, remains to be found.

(1) Amelia carried the biggest elephant over the longest distance (relative reading)
(2) Every suspect read the same book at the same time.

After presenting the problem, I will explore three solutions: 1) an extension of von Stechow (1984), 2) a solution exploiting polyadic quantification inspired in spirit by Fox & Johnson (2016), 3) a postsuppositional account following Brasoveanu (2012). I will show that the challenges faced by all three solutions are daunting, if not insurmountable. I will suggest a revision of 1) but will mainly leave the puzzle open for the audience’s insights to express themselves freely.

Speaker: Jonathan Bobaljik (Harvard University), joint work with David Koester (University of Alaska Fairbanks), Chikako Ono (Chiba University), and G. D. Zaporotskij
Title: Text setting in an Itelmen khodila (song)
Time: Wednesday (4/24), 5:30pm-6:30pm
Location: 32-D831

Abstract:
available here

Speaker: Tatiana Bondarenko & Stanislao Zompí (MIT)
Title: Leftover Agreement in Kartvelian
Time: Thursday, 4/25, 12:30-1:50pm
Location: 32-D461

Abstract: 
In this talk we argue that some instances of agreement result from probes interacting with unspelled-out features of lower agreement probes. We call such agreement Leftover Agreement, and show how it can account for the observed variation in number agreement across the four Kartvelian languages.

In Georgian and Svan, (1), the lower probe (the prefix) spells out both first-person and plural features of the object, so the higher probe (the suffix) does not find anything left over to agree with. In Laz and Megrelian, (2), however, the lower probe has spelled out only the person, but not the number feature of the object. This leftover feature is being agreed with and spelled out by a higher probe (the suffix -an / -a(n)). 
            In our talk, we will spell out our assumptions about the Kartvelian agreement system and discuss our implementation of Leftover Agreement, compare our proposal to other approaches (Halle & Marantz 1993, Harley & Lomashvili 2011, Blix 2016, Foley 2017, Thivierge 2018, a.o.), and show how it can be extended to the “inverse” agreement alignment with minimal changes.

Last Saturday evening, over 200 MIT Linguistics alums, colleagues, and friends gathered to celebrate Noam Chomsky’s spectacular 90th year. Here he is, blowing out his candles (not 90 of them, however):

Itai Bassi (4th year) presented his work on Fake Indexicals in the workshop on Dependency in Syntactic Co-variance that took place last week in Leipzig University. His handout can be found here.

Speaker: Aynat Rubinstein (The Hebrew University of Jerusalem)
Title: Desire in motion
Time: Friday, April 26, 3:30-5:00pm
Room: 32-141

Abstract:
Motion verbs are famous for their tendency to undergo language change. Across languages, verbs meaning ‘come’ and ‘go’ become future markers, aspectual markers, modals expressing necessity, and more. Tracing the development of ‘come’ in Hebrew during its revival, this talk highlights yet another diachronic pathway that stems from motion: the pathway from motion to desire. Using goal-orientation as the essential meaning component of directed motion, I offer an analysis of the internally-motivated changes in the verb’s meaning, as well as changes instigated by language contact. The investigation supports the idea that meaning change is driven not by conventionalization of pragmatic inferences but by re-distribution of semantic content in possibly innovative syntactic configurations (Beck 2012, Beck & Gergel 2015, Condoravdi & Deo 2014).

Speaker: Qi Hao (Harvard/Peking University)
Title: The Syntax/Semantics of Numeral Classifiers in Mandarin Chinese and Numeral Mapping Parameter
Time: Wednesday, April 17th, 1-2PM
Location: 32-D461

Abstract:
It is a well-known fact that classifiers are needed for grammatical counting in classifier languages such as Mandarin Chinese.

(1) a. yi ge xiaohai
one CLGENERAL child

b. san ben shu
three CLVOLUME book

c. liang ping shui
two CLBOTTLE water

There are three basic and related questions we want to ask about the Num-Cl-N Construction:

A. What is the internal constituency, [Num-Cl]-N or Num-[Cl-N]?
B. What is the function/semantics of classifiers?
C. How to account for the presence/absence of the classifier system among languages?

The mainstream analysis in generative framework is that classifiers are functional morphemes on the extended projection of nominals, functioning as divider heads (Borer 2005), type shifters from kinds to predicates (Chierchia 1998, Jiang 2012) or the syntactic counterpart of COUNT operation (Rothstein 2010, X.P. Li 2013) to make mass/kind-denoting nouns countable with numerals. That is to say, classifiers take nouns as complements and Cl-Ns further combine with numerals, structured as [Num-[Cl-N]]. And the presence of a classifier system is correlated with the absence of plural morphology as well as the absence of articles (see Chierchia 1998). However, analyses within this tradition will face difficulties when it comes to languages like Old Chinese. Old Chinese lacks both plural morphology and articles, but classifiers do not show up, as in (2).

(2) san ren xing, bi you wo shi yan. (The Analects)
three person walk must have my teacher at-there
‘If three people walk together, there must be a teacher of mine among them.’

Furthermore, it would be very hard to explain the fact for the former treatments that numerals can stand alone in predicate and argument positions in English (as in (3)) as well as in Old Chinese (as in (5)), but not in Modern Mandarin (as in (4)).

(3) a. Apples on the table are three.
b. As for apples, I bought three.

(4) a. Zhuozi-shang de pingguo shi *san/san-ge.
Table-LOCTOP MOD apple COP three/three-CLGEN. (same intended meaning as in 3a)
b. Pingguo ma, wo mai-le *san/san-ge
Apple PART, I buy-ASPT three/three-CLGEN. (same intended meaning as in 3b)

(5) a. Shi you bu ke zhi zhe san. (Records of the Grand Historian)
Thing have not can know NOM three
‘Things which cannot be known are three.’

b. Zheng yue, zuo san jun, san fen gongshi er ge you qi yi. (Zuo Zhuan)
First month, form three army, three divide country CONJ each have its one
“In the first month, (Jiwuzi) formed three armies, and divided the country into three parts and then each (of the three people) had one of them.”

Following the spirit of Semantic Parameter in Chierchia (1998), I propose that a parameter for different semantics of numerals is required to account for above facts, which can be called “Numeral Mapping Parameter”. It can be represented as follows:

(6) a. Numerals map to type in non-classifier languages such as English and Old Chinese.
b. Numerals map to type n (numbers/cardinals) in classifier languages such as Chinese.

Precisely, we should say that numerals in all languages start as individual expressions of type n denoting numbers. And the et-type use is derived from the number expression by following processes (Landman 2004, Rothstein 2009):

(7) a. three: 3 (type n)
b. IDENT(three): λn. n = 3 (raised to a predicate by IDENT)
c. function composition: λn. n = 3 °| | (composed with cardinality function)

                    = λx. |x| = 3

The derivation in (7) is a lexical rule in English, and numerals are type-et as syntactic primitives. However, Mandarin lacks such a rule in the lexicon, and hence numerals are type n when they come into syntax. Such semantic deficiency of numerals is the genuine reason why we see classifiers in Mandarin. Classifiers are of type which turns n-type numerals into et-type numerals. The semantics for classifiers can be represented as follows:

(8) a. [[Cl]] = λP λn λx [PUNIT(x) & |x|P = n] if PUNIT(x) is defined, else
[[Cl]] = λP λn λx [|x|P = n] (the case for kilos, liters)
(P stands for classifier roots)

b. [[san-Cl-ben]] = λx BENUNIT (x) & |x|BEN = 3
(paraphrase: x comes in the naturals units of volumes and the cardinality of x is three measured by the units of volumes)

Then we can say the “classifier” is a built-in semantics of English numerals (with very abstract meaning as ‘object unit’), while such a category must be independently encoded in Mandarin syntax due to Numeral Mapping Parameter. And the internal structure of the Num-Cl-N sequence is [Num-Cl]-N, contrary to most of the former treatments. Further evidence will be given to support this.

We will also make a reply to the structural ambiguity analysis for individuating readings and measuring readings of classifier construction by X.P. Li (2013) and argue that we can get the two different readings from a unified syntax. And we will make a preliminary attempt to answer the question whether the mass-count distinction really exists in classifier languages such as Mandarin.

In this week’s meeting of Phonology Circle, Anton Kukhto (MIT) will lead a discussion of Shih & Zuraw’s 2017 paper: Phonological conditions on variable adjective and noun word order in Tagalog. You can download the paper here.

Discussion leader: Anton Kukhto (MIT)
Paper: Shih, S. S., & Zuraw, K. (2017). Phonological conditions on variable adjective and noun word order in Tagalog. Language, 93(4), 317-352 (available here).
Time: Wednesday (4/17), 5:00-6:30pm
Location: 32-D831

Abstract:
Tagalog adjectives and nouns variably occur in two word orders, separated by an intermediary linker: adjective-linker-noun versus noun-linker-adjective. The linker has two phonologically conditioned surface forms, -ng and na. This article presents a large-scale corpus study of adjective/ noun order variation in Tagalog, focusing in particular on phonological conditions. Results show that word-order variation in adjective/noun pairs optimizes for phonological structure, abiding by phonotactic, syllabic, and morphophonological well-formedness preferences that are also found elsewhere in Tagalog grammar. The results indicate that surface phonological information is accessible for word-order choice.

Speaker: Conor McDonough Quinn (University of Southern Maine)
Title: Animacy, obviation, inverse, and delightful phonological subtleties: a call to look more at Passamaquoddy-Wolastoqew(Maliseet) and relatives, and helps for formal theory from endangered language pedagogy
Time: Thursday, 4/18, 12:30-1:50pm
Location: 32-D461

Abstract:

Starting with core discussions of three traditionally thorny phenomena Algonquian morphosyntax:

• Algonquian grammatical animate status as formal “mission creep” from semantic animacy, via highly constrained analogical “families” that shift over time and space, but are reliably productive for any given speaker.
• Proximate-obviative as marking referential entailment dependency: in 3rd+3rd Goal-Theme configurations (verbal-ditransitive or nominal-possessive), the Theme cannot be proximate relative to the obviative Goal; only the reverse. E.g. in [HER MOTHER], [MOTHER] must always be obviative. This restriction seems tied crucially to the observation that knowing the complete reference of [MOTHER] in [HER MOTHER] entails knowing the reference of HER.
• Inverse for 3→{1,2} configurations in Algonquian languages being only strictly required for clause-types realized morphosyntactically as possessed nominals: in them, core transitive arguments are introduced via a Goal-Theme (Possessor-Possessee) configuration, and so are subject to a Person-Case Constraint effect.

We then briefly note three other features in desperate need of in-depth research: (a) head-marking for oblique (spatial, manner, temporal, etc.) arguments; (b) nominal tense; (c) standalone Secondary Objects (= morphosyntactically same as ditransitive Themes, but with no overt/interpreted Goal argument). Also briefly sketched are three under-researched phonetic-phonological phenomena: (a) iambic weak/strong-schwa alternations + related rich initial/final consonant clusters; (b) contrastive pitch-accent + final vowel deletion alternations; (c) preaspiration and gemination in a voicing-noncontrasting phonational system. Finally, we observe how being “reduced to” minimalist, non-technical, and pragmatic-communicatively-grounded presentations of these kinds of unfamiliar linguistic phenomena—namely, doing what it takes to teach them genuinely effectively to adult learners—creates fertile ground for innovative rethinking.

A longer version of the abstract can be found here.

Speaker: Tal Linzen (Johns Hopkins University)
Title: Linguistics in the age of deep learning
Time: Thursday, 4/18, 5-6pm
Location: 32-141

Abstract:

Deep learning systems with minimal or no explicit linguistic structure have recently proved to be surprisingly successful in language technologies. What, then, is the role of linguistics in language technologies in the deep learning age? I will argue that the widespread use of these “black box” models provides an opportunity for a new type of contribution: characterizing the desired behavior of the system along interpretable axes of generalization from the training set, and identifying the areas in which the system falls short of that standard.

I will illustrate this approach in word prediction (language models) and natural language inference. I will show that recurrent neural network language models are able to process many syntactic dependencies in typical sentences with considerable success, but when evaluated on carefully controlled materials, their error rate increases sharply. Perhaps more strikingly, neural inference systems (including ones based on the widely popular BERT model), which appear to be quite accurate according to the standard evaluation criteria used in the NLP community, perform very poorly in controlled experiments; for example, they universally infer from “the judge chastised the lawyer” that “the lawyer chastised the judge”. Finally, if time permits, I will show how neural network models can be used to address classic questions in linguistics, in particular by providing a platform for testing for the necessity and sufficiency of explicit structural biases in the acquisition of syntactic transformations.

Speaker: Dan Lassiter (Stanford University)
Title: Mathematical Counterfactuals
Time and Place: Friday, April 19, 3:30-5:00pm, room 32-141
Abstract:

Counterfactual reasoning about mathematical truths (“If 7 + 5 were 11, I’d have gotten a perfect score on the math test”) presents an important challenge to standard accounts of the semantics of conditionals. I describe a semantics based on Pearl-style interventions on generative models and show that it provides a simple account of mathematical counterfactuals that also coheres well with research on mathematical cognition. The approach is related to the possible worlds theory in that the models are recipes for generating descriptions of possible worlds, but their procedural character is crucial in supporting interventions and the kind of partiality that I argue we need to render mathematical counterfactuals meaningful.
_______________________________________________

Speaker: Danfeng Wu (MIT)
Title: Prefer the less specified form: Evidence from Lebanese Arabic
Time: Tuesday, 4/9, 1-2pm
Location: 32-D461

Abstract:
For languages with resumptive pronouns (RPs), an Economy principle has been proposed that prefers gap realization whenever possible (e.g. Shlonsky 1992, Pesetsky 1998, McDaniel and Cowart 1999, Sichel 2014, and Rasin 2016). In Hebrew, for instance, the A’-trace in a raising relative clause is realized as a gap, unless an island blocks the movement, in which case an RP is inserted in the trace position. Because an RP has more internal structure than a gap, this Economy principle can then be stated either as a preference for the leastspecified form or the preference for the less specified form.

In this preliminary and informal discussion I argue that the latter formulation of the Economy principle is correct. I present evidence from Lebanese Arabic, which has a richer resumption strategy than Hebrew and can use an independent morpheme (“strong” pronouns), a clitic (“weak” pronoun), or an epithet in resumption (Aoun et al. 2001). The strong pronoun and the epithet have more morphological complexity than the weak pronoun. When coindexed with a quantificational antecedent (a wh-phrase or a quantifier), the strong pronoun and the epithet have the same distribution on the one hand, while the weak pronoun and the gap pattern together and have a wider distribution on the other hand. Strikingly, a weak pronoun alternates with a wh-gap in non-island contexts, and the wh-antecedent can be reconstructed to the position of the weak pronoun (Aoun and Benmamoun 1998). Based on these facts I argue that the Economy principle is a comparative one rather than a superlative one, preferring the less specified form whenever possible. In the case of Lebanese Arabic, it prefers a gap and a weak pronoun rather than a strong pronoun and an epithet.

Speaker: Christopher Baron (MIT)
Title: Entailments, implicatures, and absolute adjectives
Time: Wednesday, April 10th, 1-2PM
Location: 32-D461

Abstract: 
Absolute adjectives like straight and bent give rise to interesting entailments and implicatures when they occur in degree constructions; the [a] and [b] examples below are comparatives, and the [c] examples are degree achievements.

1. [a] Bar A is straighter than it was before.
   [b] Bar A is straighter than Bar B is.
   [c] Bar A straightened.
2. [a] Bar C is more bent than it was before.
   [b] Bar C is more bent than Bar B is.
   [c] Bar C bent

(1a) and (1c) entail that Bar A wasn’t completely straight before; (1b) entail that Bar B isn’t completely straight. However, (1a) and (1b) seem to imply that Bar A isn’t completely straight, whereas (1c) seems to imply that it is completely straight (now). The examples in (2) are similar. All three entail that Bar C is bent (now). (2a) seems to imply that Bar C was already bent, and (2b) seems to imply that Bar B is also bent. However, (2c) seems to imply that Bar C wasn’t bent before. 

I’ll argue that this implied content really is implicature, rather than entailment or presupposition, and explore and expand on accounts of some (but not all) of these implicatures (e.g. Kennedy 2007). Furthermore, I’ll explore various possibilities for the apparent reversal in implicature with degree achievements, and argue that none of the obvious solutions are quite appealing. I leave open for now, however, a positive account of these data. 

Speaker: Jonathan Bobaljik (Harvard University), joint work with David Koester (University of Alaska Fairbanks), Chikako Ono (Chiba University), and G. D. Zaporotskij
Title: Text setting in an Itelmen khodila (song)
Time: Wednesday 10th, 5:00pm-6:30pm
Location: 32-D831

Abstract: available here

Presenter: Yadav Gowda (MIT Linguistics)
Paper to read: Tanenhaus et. al (1995)
Time: Thursday, 4/11, 5-6pm
Location: 46-5165

Details: 
Tanenhaus et. al 1995 is an influential paper arguing that language processing is (a) incremental and (b) provides us with an argument against the modularity of syntax. I will review their arguments and explore related questions, including, but not limited to: What defines a module? Why should/shouldn’t cognitive scientists propose modules? What should we take to be neurological/psychological evidence of a module?

Noam Chomsky will be giving two lectures at MIT this week.

Location : 37-212
Times : Wednesday 10th of April & Friday 12th of April

Hoping to see you there!

Our alum Newell Lewey has accepted a position of adjunct instructor at the University of Maine at Machias, where he will teach Intro to Passamaquoddy and Public Speaking. Congratulations, Newell! Kuli-kiseht!

Norvin Richards spent April 5-6 at the Princeton Symposium on Syntactic Theory (PSST), as did recent alumna Michelle Yuan. The theme of this year’s meeting was “counterexamples”. Norvin reports the following feeling:
“[I’m] trying not too think too hard about the fact that, when the organizers tried to think of people whose theories have lots of counterexamples, they apparently thought of [me] right away.”

We are absolutely thrilled to announce that Michelle Yuan (PhD 2018), who received her PhD from our department last summer, has accepted a tenure-track Assistant Professor position in the Department of Linguistics at the University of California at San Diego. Also joining UCSD faculty is Emily Clem, who is currently finishing at Berkeley and was a visitor at MIT in Spring 2018. Congratulations to both!

Speaker: Rafael Abramovitz and Itai Bassi (MIT)
Title: Relativized Anaphor Agreement Effect: Evidence from Koryak
Time: Tuesday, 4/2, 1-2pm
Location: 32-D461

Abstract:
In this talk, we discuss the Anaphor Agreement Effect (AAE), an observation first proposed in Rizzi (1990) to the effect that anaphors do not trigger phi-agreement. We report on novel data from Koryak, a Chukotko-Kamchatkan language of the Russian Far East, which we claim constitutes a particularly revealing partial counterexample to previous formulations of the AAE. Specifically, we show that the Koryak anaphor uvik triggers obligatory number agreement consistent with its binder, but never triggers person agreement consistent with it. We therefore propose that the AAE is limited to preventing anaphors from triggering covarying person agreement, but that it does not block number (or gender) agreement. Our account of this combines arguments from recent work by Omer Preminger on the AAE, as well converging evidence about the syntax of phi-features from Moskal (2015), Harbour (2016), and van Urk (2018). Following Preminger (2019), we take the AAE to be the result of the way that anaphors are structured: they contain a head Anaph that blocks agreement into its complement. However, to account for the Koryak facts (as well as other partial counterexamples to the AAE discussed in Murugesan (2018)), we deviate from Preminger’s proposal by allowing languages to vary with respect to where Anaph merges along the pronominal spine. In languages like Koryak, where anaphors trigger number agreement, Anaph merges below Num, whereas in languages like Albanian, where they trigger only 3SG agreement (Woolford 1999, Run Chen p.c.), it merges above Num. Following Moskal (2015) et. seq., we take person to be the most embedded phi-feature, forcing agreement with it to be blocked by Anaph regardless of how low it merges. Finally, we propose a compositional semantics for the Anaph head that explains how its presence gives rise to co-reference between the two arguments of the relevant predicate.

We’re very happy to announce that Simon Charlow will be here this week, and will teach a mini-course Wednesday and Thursday, April 3rd and April 4th. Details are below. 

Speaker: Simon Charlow (Rutgers)
Title: Extensible semantics
Time: Wednesday (4/3), 1:00-2:30pm  and Thursday (4/4), 3:30-5:00pm
Place: 32-D461 on Wednesday, 37-212 on Thursday

Abstract: Since the basics were outlined by Frege, semantic theory has been enriched in various ways. Pronouns and alternative-invoking expressions motivated a souped-up interpretation function, one capable of composing assignment-sensitive and multifarious meanings (and eventually meanings that were both). Various kinds of ‘dynamic’ phenomena resulted in lexical enrichments that replaced propositions with context-change potentials. Quantification and scope gave us the Y-model. Type-shifting has appeared from time to time.
This mini-course will show how to recast all these enrichments in a uniform way with basic, well-studied tools used by programmers to introduce so-called ‘side-effects’ into functional programs: (applicative) functors and monads. We’ll see concrete examples of how to abstract these tools out of existing semantic theories, and compositionally combine them to build new super-tools (in more ways than one). We’ll study how this flexible, modular approach to composition solves various empirical problems and dissolves some recalcitrant technical issues. And we’ll apply these techniques to the study of pronouns (with or without assignments), exceptionally scoping indefinites, cross-sentential and donkey anaphora, quantification and scope (with and without events), and ‘higher-order’ analogs of these phenomena (e.g., functional pronouns and higher-order questions).

Suggested readings: [1] “The scope of alternatives“ And optionally, one of the following: [2] ”A modular theory of pronouns and binding“[3] “Variable-free semantics and flexible grammars for anaphora”

Speaker: Daniel Hole (Universität Stuttgart)
Title: Arguments for a universal distributed syntax of evaluation, scalarity and basic focus quantification with ‘only’
Time: Thursday, 4/4, 12:30-1:50pm
Location: 32-D461

Abstract: In this talk, I review the evidence that has been adduced for a multi-constituent syntax of focus particle constructions. Traditionally, those components that I model as independent morphemes with their own scope-taking properties have been analyzed as submorphemic components of focus particles. I use ‘only’ words to make this point. This work is based on Hole (2013, 2015, 2017), and it makes use of data from Chinese, Vietnamese, German and Dutch. However, many arguments carry over to English. Time allowing, I will also present novel data from the interaction of German nur with modals and the German NPI modal brauchen ‘need (+NPI)’.

This approach to focus particles stands in stark contrast to Büring & Hartmann (2001) or Coppock & Beaver (2013).

Speaker:  Candace Ross (MIT CSAIL)
Title: Grounding Language Acquisition by Training Semantic Parsers using Captioned Videos
Time: Thursday, 4/4, 5-6pm
Location: 46-5165

Abstract:
We develop a semantic parser that is trained in a grounded setting using pairs of videos captioned with sentences. This setting is both data-efficient, requiring little annotation, and similar to the experience of children where they observe their environment and listen to speakers. The semantic parser recovers the meaning of English sentences despite not having access to any annotated sentences. It does so despite the ambiguity inherent in vision where a sentence may refer to any combination of objects, object properties, relations or actions taken by any agent in a video. For this task, we collected a new dataset for grounded language acquisition. Learning a grounded semantic parser — turning sentences into logical forms using captioned videos — can significantly expand the range of data that parsers can be trained on, lower the effort of training a semantic parser, and ultimately lead to a better understanding of child language acquisition.

Speaker: Simon Charlow (Rutgers)
Title: Local contexts in ellipsis
Time: Friday, April 5th, 4pm-5:30pm
Location: 32-155

Abstract: This talk advocates a specific semantic implementation of discourse congruence (Rooth 1992, Schwarzschild 1999), which plays a central role in the licensing of ‘anaphoric’ reduction (i.e., ellipsis and deaccenting). I argue that congruence is checked compositionally, and that satisfaction of congruence is determined extensionally — that is, in a local context. I show how this leads to significant simplifications in the theory of reduction licensing, allowing us to dispense with otherwise necessary prohibitions on Meaningless Coindexing (Sag 1976, Heim 1997) and Redundancy within an assignment (cf. Schlenker 2005), and to strengthen the relationship between an elided phrase and its antecedent to one of pure identity.
I consider several consequences of my proposal for the formulation of congruence operators, arguing that it compels us to take their anaphoric character seriously, while allowing for ex post facto (i.e., post-suppositional) linking of congruence operators to their ‘antecedents’. And I explore some consequences of these moves for restrictions on antecedent-contained deletion (Kennedy 1994) and the puzzling phenomenon of focused bound pronouns (Sauerland 1998).

The 15th workshop on Altaic formal linguistics will take place at Moscow State University on September 26th to 28th. The deadline for abstract submissions is April 14th 2019.

From April 2 to April 5, Arts at MIT, in collaboration with Prof. Michel DeGraff at MIT Linguistics and the MIT-Haiti Initiative, and Prof. Nick Montfort in Comparative Media Studies, is hosting the visit from Haiti of rapper, poet and singer BIC Tizon Dife (Roosevelt Saillant) for a celebration of music, literature, linguistics, dance and community. Joining BIC at his concert on Friday, April 5, 8pm, at MIT, is the Boston-based dance company Jean Appolon Expressions (JAE). JAE will also offer a dance workshop on Friday, April 5, 1-3pm. BIC will help lead a digital story-telling workshop on Tuesday, April 2nd, 5-7pm, and he will perform at Brothers’ Kafe Kreyòl in Everett on Wednesday, April 3rd, 8pm. All events are free of charge and open to the MIT and greater Boston community. For more information on BIC’s schedule and related events, please visit https://arts.mit.edu/artists/bic/#schedule.

Second year student, Sherry Yong Chen, presented two projects at the 32nd CUNY Annual Conference on Human Sentence Processing:

1. “QUD effects on scope ambiguity involving comparative quantifies” - Sherry Yong Chen, Leo Rosenstein and Martin Hackl
2. “Interference effects in the memory retrieval of presuppositional dependency” - Sherry Yong Chen and E. Matthew Husband

In other news, her co-authored chapter has just appeared in the Oxford Handbook of Experimental Semantics and Pragmatics (https://global.oup.com/academic/product/the-oxford-handbook-of-experimental-semantics-and-pragmatics-9780198791768): Chapter 5 “Event (de)composition”, Sherry Yong Chen and E. Matthew Husband. 

Congratulations to our alum Ted Levin (PhD 2015) on his new job at Facebook in Seattle! Here’s what he writes to us about his new position:

“I’ve been working as a Linguist at Facebook Reality Labs for the past month or so. I’ve found the work to be challenging with lots of new things to learn, and am enjoying it despite it being quite different than my academic work. Broadly speaking, the team of ~20 linguists that I’m a part of sits at the very beginning of a pipeline to make Facebook apps better at understanding user intentions through voice or text.

“Narrowly, my job is to identify ways in which users might want to interact with Facebook apps (e.g. send messages, get recommendations, set reminders), collect data to exemplify the utterances that satisfy requests within these domains, and oversee the annotation of this data so that it can be utilized to train machine learning algorithms farther down the pipeline.”

News has reached us from our recent alum Wataru Uegaki (PhD 2015) that he has accepted a position as a Lecturer in Semantics (≅ Assistant Professor) in the Linguistics department of the University of Edinburgh. Wataru is currently an Assistant Professor at the University of Leiden, in the Netherlands. Congratulations, Wataru!!

Third-year student Yadav Gowda spoke on “Movement within and without a clause” at Formal Approaches to South Asian Languages (FASAL9) this weekend at Reed College.

Speaker: Itai Bassi (MIT)
Title: Principle C effects with plural antecedents (joint work with Paul Marty)
Time: Monday, March 18th, 1-2 pm
Abstract:In the literature on Principle C, it has been claimed on the basis of examples like (1) that an R-expression cannot overlap in reference with a c-commanding plural antecedent (a.o., Lasnik 1989, Schlenker 2003). Yet we add to (1) the datapoint in (2), and observe that this claim only holds on a distributive interpretation of the relevant antecedent: as (2) shows, things get better on a collective interpretation (cf. Reinhart and Reuland for a parallel observation for principle B configurations).

(1) They think that Oscar should be the band leader (* if Oscar is part of the reference of “they”)
(2) They (jointly) decided that Oscar will be the band leader (ok if Oscar is part of the reference of “they”)

Our idea to understand this contrast is this: the distributive LF, but not the collective one, has an embedded constituent of the form ‘x thinks that Oscar…’, where ‘x’ is an atomic variable quantified over by a distributive operator.  When `x’ takes Oscar as one of its values, a (classical) Principle C violation obtains. We propose to derive these “embedded principle C” effects by extending Reinhart’s competition theory of principle C (aka ‘Rule I’) to make it operative at embedded levels. We then argue that on this refinement, Rule I can be eliminated as a primitive of the grammar and be subsumed by the independently motivated principle, Maximize Presupposition!, which has also been argued to regulate competition between LF representations at embedded levels (Singh, 2011).

Speakers: Colin Davis (MIT)
Title: Mismatched suppletion in Azeri as morphology/phonology competition
Time: Monday, March 18th, 5-6:30pm
Location: 32-D831

Abstract: (longer PDF version available at http://tinyurl.com/colin-morphun)

In this paper, I examine a puzzle about suppletion in the northern dialect of Azeri/Azerbaijani (Turkic). I focus on suppletion of the perfect (PRF) / evidential (EVID) morpheme, whose default form is -miʃ. My fieldwork work has found, in agreement with Oztopcu (2003), that this morpheme has an allomorph -Ib that is typically only used with 2nd and 3rd person subjects:
(1)    a.  o      gatʃ-mɨʃ/ɨb                                     
             1SG run-PRF-3SG                                            
             ’He/she/it has run’        

        b.  biz   dʒal-miʃ/*ib-ij   
             1PL come-PRF-1PL                              
             ‘We have come’  

However, I have found that in contexts where multiple adjacent instances of -miʃ would surface, one of those instances is realized as -Ib, even when the subject is 1st person:

(2)    a.  biz  gatʃ-ɨb-mɨʃ-1ɣ                                                   1PL run-PRF-EVID-1PL                                      
            ’Apparently we had run’                                       

         b.  man je-maj       je-ib-miʃ-am  
              1SG eat-NMLZ eat-EVID-PRF-1SG  
             ‘Apparently I ate food’  

Since -Ib suppletion typically requires a 2nd/3rd person subject, we would have expected the 1st person subjects in (2) to make -Ib unavailable. Why did -Ib suppletion in (2) succeed? I argue that this over-application of suppletion in (2) occurs due to a phonological constraint against forms with adjacent identical morphemes (Menn & MacWhinney 1984, Plag 1998, Yip 1998, a.o.). This phonologically-forced morphological mismatch provides new evidence that morpheme insertion interacts and competes with phonological constraints (Wolf 2008, 2009, Pertsova 2015).

Speaker: Cater Chen
Title:  Split Partitivity in Mandarin
Time: Tuesday, March 19, 1pm-2pm
Location: 32-D461

Abstract: In Mandarin (Chinese), a numeral classifier (henceforth NCL) fragment can be interpreted as a partitive expression relative to a definite DP antecedent in various constructions, a phenomenon I refer to as split partitivity (henceforth SP). In the first part of this talk, I will argue for a stranding approach to SP. That is, the NCL fragment is analyzed as a stranded element in the course of movement of its nominal associate, therefore, it can be used to identify a movement trace of its nominal associate. 

In Mandarin (Chinese), a numeral classifier (henceforth NCL) fragment can be interpreted as a partitive expression relative to a definite DP antecedent in various constructions, a phenomenon I refer to as split partitivity (henceforth SP). In the first part of this talk, I will argue for a stranding approach to SP. That is, the NCL fragment is analyzed as a stranded element in the course of movement of its nominal associate, therefore, it can be used to identify a movement trace of its nominal associate. 

(1)             NCL fragment identifies A-traces

a.  [Na-liu-ge     xuesheng]_i  [_WollP  hui you  [_vP   [_NCLP  san-ge t_i] lianxi     Zhangsan]].

      Dem-6-CL     student                   will exist                   3-CL          contact Zhangsan

Lit. ‘Those six students will three (of them) contact Zhangsan.’

b.  Na-liu-ge   xuesheng [_VP lai-le          san-ge].             

Dem-6-CL student           come-Perf 3-CL

‘Three of those six students came.’

Lit. ‘Those six students came three (of them).’

(2)       NCL fragment identifies Ā-traces

a.  [Na-liu-ge   xuesheng]_i (a),   you [_NCLP san-ge t_i] renshi  Zhangsan.                     

Dem-6-CL student        Top  exist          3-CL         know   Zhangsan    

‘Those six students, three (of them) know Zhangsan.’

b.  [Na-liu-ge  xuesheng]_i (a),  Zhangsan renshi  [_NCLP san-ge t_i].              

Dem-6-CL student       Top Zhangsan know              3-CL

‘Those six students, Zhangsan knows three (of them).’

In the second part of this talk, I will show that, having justified a stranding approach to SP, we can use SP as a tool to study various constructions in Mandarin that involve argument-gap dependencies. In particular, passive constructions in Mandarin have been analyzed as involving raising, control or null operator movement and predication. I propose that SP can shed light on the right analysis, because different patterns of SP can diagnose different types of argument-gap dependencies.

Speaker: Adam Albright (MIT), joint work with Vighnesh Subramaniam (Millard North High School, Omaha NE)
Title:  Modeling typological frequency with a grammatical learner
Time: Wednesday (3/20), 5:00pm-6:30pm
Location: 32-D831

Abstract: available here

Speaker: Philip Shushurin (NYU)
Title: A uniform account of possessives and applicatives: evidence from external possession in Russian
Time: Thursday, 3/21, 12:30-1:50pm
Location: 32-D461
Abstract: Much of the work on external, or raised, possessors in Russian (Paykin and van Peteghem (2003), Grashchenkov and Markman (2009)), as well as in other languages with the similar phenomenon (Landau (1999), Deal (2016)), has recognized the dual nature of such arguments: on one hand, they are interpreted as possessors, on the other hand, they show many similarities with other types of arguments, most frequently, applicatives and topics. I consider two external possession constructions in Russian and propose that they are merged in a DP-external functional projection (ApplP) and either are licensed in situ or move to a topic position for licensing. I propose that goals of ditransitives (low applicatives), external possessors and DP-internal possessors are introduced by the same functional head.

Speakers: Anya Lunden (William & Mary)
Title: Real and accidental glides
Time: Friday, March 22nd, 3:30pm-5pm
Location: 32-141

Abstract:

Articulating a vowel sequence like [i.a] results in formant movements that are identical to those resulting from the articulation of a [j]. This raises questions about the status of phonological rules of glide insertion, which are often stated as the hiatus resolution strategy for such configurations. The talk explores the question of whether there are, or can be, “real” glides in such cases, and whether the “accidental” glides have any status perceptually. Evidence from production and perception studies as well as cross-linguistic behavior is presented and used to motivate an analysis where an accidental fact about articulation has consequences for the phonology.

Speaker:  Paola Merlo (University of Geneva)
Title: Quantitative Computational Syntax: some case studies
Time: Thursday, 3/21, 5-6pm
Location: 46-5165
Abstract: In the computational study of intelligent behaviour, the domain of language is distinguished by the complexity of the representations and the sophistication of the domain theory that is available. It also has a large amount of observational data available for many languages. The main scientific challenge for computational approaches to language is the creation of theories and methods that fruitfully combine large-scale, corpus-based approaches with the linguistic depth of more theoretical methods. I report here on some recent and current work on word order universals and argument structure that exemplifies the quantitative computational syntax approach. First, we demonstrate that typological frequencies of noun phrase orderings, universal 20, are systematically correlated to abstract syntactic principles at work in structure building and movement. Then, we investigate higher level structural principles of efficiency and complexity. In a large-scale, computational study, we confirm a trend towards minimization of the distance between words, in time and across languages. In the third case study, much like the comparative method in linguistics, cross-lingual corpus investigations take advantage of any corresponding annotation or linguistic knowledge across languages. We show that corpus data and typological data involving the causative alternation exhibit interesting correlations explained by the notion of spontaneity of an event. Finally, time permitting, I will discuss current work investigating on whether the notion of similarity in the intervention theory of locality is related to current notions of similarity in word embedding space.

Bio:
Paola Merlo is faculty in the linguistics department of the University of Geneva. She is the head of the interdisciplinary research group Computational Learning and Computational Linguistics (CLCL). The group is concerned with interdisciplinary research combining linguistic modelling with machine learning techniques. The scope of her current research includes fundamental issues in the statistical nature of language, empirical evaluations of linguistic proposal about the lexical semantics of verbs and language universals of word order and statistical models of syntactic and semantic parsing. Prof. Merlo has been the editor of the journal of the Association for Computational Linguistics, Computational Linguistics, and has been member of the executive committee of the EACL and of the ACL. Prof. Merlo holds a doctorate in Computational Linguistics from the University of Maryland, USA. She has been associate research fellow at the Institute for Cognitive Science at the University of Pennsylvania, and has been visiting scholar at Rutgers, Edinburgh, and Stanford.

Title: How to find the rabbit in the big(ger) box:
Reasoning about contextual parameters for gradable adjectives under embedding (joint work with Roger Levy (MIT BCS) and Elizabeth Coppock (Boston University))
Speaker: Helena Aparicio (MIT BCS)
Time: Friday, 22nd of March 2-3pm
Abstract:Haddock (1987) noticed that the definite description ‘the rabbit in the hat’ succeeds in referring even in the presence of multiple hats, so long as only one hat contains a rabbit. These complex definites suggest that uniqueness with respect to the NP hat is not required in such embedded contexts, raising the question of what the correct formulation of the uniqueness condition for definite determines is. Generally speaking, two types of solutions have been proposed to this puzzle. The first one postulates a complex semantic representation for definite determiners, where uniqueness can be checked at different points of the semantic representation for either sets of hats or sets of rabbit-containing hats (Bumford 2017). The second type of account proposes that definite descriptions can be evaluated against a sub-portion of the maximally available context (Evans 2005; Frazier 2008; Muhlstein 2015). This pragmatic mechanism ensures that reference resolution is successful, even when the maximal context would violate the uniqueness presupposition of the definite article.

The present work seeks to tease apart these two classes of theories by investigating the interpretive preferences for similarly embedded noun phrases containing a positive or comparative adjective (e.g., ‘the rabbit in the big/ger box’). Experimental results show that embedded positive adjectives exhibit a sensitivity to contextual manipulations that embedded comparatives lack. We derive this sensitivity using a probabilistic computational model of the contextual parameters guiding the interpretation of the embedded NP, and compare it to alternative models that vary in the lexical representations assumed for definite determiners. Our simulation results show that neither of the two proposals under consideration can independently account for all of the observed experimental results. We show that the model that best matches human data is one that combines both a complex uniqueness check (à la Bumford) with pragmatic context coordination.

Speakers: Suzana Fong (MIT)
Title: Regular and athematic participles in Brazilian Portuguese
Time: Monday, March 11th, 5-6:30 pm
Location: 32-D831

Abstract:

In Brazilian Portuguese, the regular participial (PTC) form of the verb can be parsed as follows:

(1) com-i-d-as
eat-TH-PTC-AGR(fem.pl)
‘eaten’ [where ‘TH’ is a theme vowel]

Besides a regular form (2a), some verbs also allow for an additional short form of the PTC (2b):

(2) a. Regular participle
limp-a-d-as
clean-TH-PTC-AGR(fem.pl)
‘cleaned’
b. Short/Athematic participle
limp-as
clean-AGR(fem.pl)
‘cleaned’

Short PTC’s like (2b) lack both the participial morpheme (-d in (1) and (2a)) and a theme vowel (-a in (2a)), hence why they are referred to as athematic PTC’s.

In this presentation, I focus on the distribution of the PTC when both a regular and an athematic form is in principle available. In particular, we will look at the PTC form(s) that occur(s) in three constructions: verbal passives (3a), adjectival passives (3b), and absolute participles (3c).
(3) a. Verbal passive: regular or athematic PTC allowed
As gavetas foram limp-a-d-as / limp-as pelo João.
the drawers were cleaned(REG) / cleaned(ATHEM) by.the João
‘The drawers were cleaned by João.’

b. Adjectival passive: only athematic PTC allowedAs gavetas estão/permanecem *limp-a-d-as / limp-as.
the drawers are/remain *cleaned(REG) / cleaned(ATHEM)
‘The drawers are/remain clean.’
c.Absolute participles: only regular PTC allowed[ Limp-a-d-as / *Limp-as as gavetas ], o João pôde ir descansar.
[ cleaned(REG) / *cleaned(ATHEM) the drawers ] the João could go rest
‘The drawers having been cleaned, João could go rest.’
I propose a tentative analysis where the athematic PTC (2b) arises as a consequence of a fusion rule that targets v and the root. Whether or not that rule can be applied will depend on the size of a given PTC construction, which in turn determines how much structure is Spelled-Out. Specifically:

• In absolute participles like (3c), v and the root are Spelled-Out separately, bleeding the context of application of the proposed fusion rule. Only a regular PTC can then be realized.
• In adjectival participles like (3b), v and the root are part of the same Spell-Out domain, which allows for fusion to apply. As a consequence, an athematic PTC can be used in this construction.
• In verbal passives like (3a), two derivations are possible. In one of these derivations, v and the root are Spelled-Out separately, just as in absolute participles, hence why a regular PTC arises. In the other — equally convergent — derivation, v and the root are Spelled-Out together, just as in adjectival passives, hence why an athematic PTC arises.

Speakers: Frank Staniszewski (MIT)
Title: Marty & Romoli (2019)
Time: Wednesday, March 13th, 1-2 pm
Location: 32-D461

Abstract:
I will present Marty & Romoli’s new paper “Presupposed Free Choice” (which expands on material discussed in  Gajewski & Sharvit (2012) and  Spector & Sudo (2017)). I will also explore possible connections with some of my work in progress on ‘until’ phrases, which I hypothesize should be treated as free choice items.

Speaker: Chris Tancredi (Keio University)
Title: De dicto, de re and de qualitate unified
Time: Thursday, 3/14, 12:30-1:50pm
Location: 32-D461

Abstract:

Past approaches to the semantics of belief statements have argued for a multiplicity of distinct interpretations, including de dicto, de re, de qualitate and de translato.  The need for ambiguity in attitude statements is clear from the potential truth of sentences like Ralph believes Ortcutt is a spy, but he doesn’t believe ORTCUTT is a spy.  However, I argue that the only ambiguity specific to attitude statements is the de translato/non-de translato distinction.  In particular, I show how to reduce de dicto/de re/de qualitate interpretations to a single form.  The key to the reduction is to analyze the embedded clause of an attitude statement as denoting a proposition inferable from an underlying belief of the subject rather than denoting the subject’s underlying belief itself.  I show that the semantics developed can account for attitudes toward necessary as well as impossible propositions, and that it further can account for the range of entailments felt to hold among multiple attitude statements.

Speaker: Danfeng Wu (MIT Linguistics)
Title: Syntactic Theory: A Formal Introduction Chapters 9.3-9.9 by Sag, Wasow and Bender (2003)
Location: 46-5165
Time: Thursday, 3/14, 5-6pm

Abstract: 

What is the role of psycholinguistic evidence (specifically evidence from language processing) in the study of language? What is the relation between knowledge of language and use of language? We hope to explore these questions through a discussion of an HPSG textbook chapter. HPSG (Head-driven phrase structure grammar) is a different syntactic framework from generative transformational grammar, and is surface-oriented, constraint-based and strongly lexicalist. This textbook chapter argues that HPSG is more compatible than transformational grammar with observed facts about language processing. For instance, language processing is incremental and rapid (e.g. Tanenhaus et al. 1995 & 1996, Arnold et al. 2002). The order of presentation of the words largely determines the order of the listener’s mental operations in comprehending them. And lexical choices have a substantial influence on processing (MacDonald et al. 1994). For these reasons, such psycholinguistic evidence supports an HPSG type of grammar, and poses difficulty to transformational grammar.

Speakers: Paula Fenger (University of Connecticut/Harvard University)
Title: Words within words: The internal syntax of verbs
Time: Monday, March 4th, 5-6:30 pm
Location: 32-D831

Abstract: Superficially, languages differ in how grammatical categories like tense and aspect are expressed on the verb; e.g. in English, both [progressive] and [past] cannot be suffixes on the verb (*stay-ing-ed) and require a periphrastic realization (was stay-ing), whereas in Turkish both [progressive] and [past] appear as agglutinating suffixes on the verb (kal-ıy’or-dı, stay-prog-past). Moreover, a language like Turkish is not uniformly agglutinating and can have periphrastic constructions in other environments, which makes it seem like the existence of periphrasis is contextually defined within a language. In this talk I explore the idea that Turkish-type languages and English-type languages are underlyingly the same. By looking at stress assignment patterns and behavior under coordination, it is revealed that the agglutinative pattern is only superficial. That is, syntax-phonology mismatches in Turkish in fact reveal a domain inside the verb and this domain is similar to the domain found for the periphrastic construction in English. Specifically, the syntactic operation that creates words (main verbs and auxiliaries) is uniform in both types of languages, the differences between them arise only at PF. This approach does not treat notions like “agglutinating” and “periphrastic” as deep properties of grammar, but as superficial differences.

Speaker: Yadav Gowda (MIT)
Title: Moving Clauses
Time: Tuesday, March 5, 1pm-2pm
Location: 32-D461

Abstract:
The first part of this talk will focus on the derivation and interpretation of so-called ‘Comp-internal clauses’ in Bengali (Bhattacharya 2001). Clauses headed by the head-initial complementizer je normally appear obligatorily extraposed to the right of the (normally head-final) verb:
(1) ami bollam [je Aparṇa deri korlo]      I       said      C Aparna late did     I said that Aparna was late.
However, when an element appears in first position within the clause (henceforth the PJE, or pre-je-element), these clauses obligatorily appear to the left of the verb, in the normal object position:
(2) ami [Aparna je deri korlo] bollam.      I       aparna  C  late did      saidI said that Aparna was late.
I will present previously unnoticed facts about the binding and scope behavior of Comp-internal clauses which suggest that the PJE occupies its own matrix specifier, despite the fact that it appears to move along with the rest of the clause. I will argue that this mismatch suggests that Comp-internal clauses involve clausal pied-piping driven by long scrambling of the PJE.
In the second part of this talk, I will consider facts which suggest that clausal scrambling in other languages (Basque, Bavarian German, Kannada) may also be amenable to an analysis in which this movement is driven by a clause-internal element, and discuss how this supports previous arguments on the immobility of CPs (Koster 1978, Stowell 1981, and many others).

Speaker: Laura Kalin (Princeton University)
Title: Morphological opacity in Turoyo (Neo-Aramaic): Support for cyclicity and separation
Time: Friday, March 8th, 3:30-5 pm
Location: 32-141 (please note, this is a different lecture hall than the room the colloquia are usually in)

Abstract: In this talk, I use complex verbs in the Neo-Aramaic language Turoyo (Jastrow 1993) as a window into theoretical issues at the syntax/morphology and morphology/phonology interfaces, including cyclicity, derivational timing, and constraints on allomorphy.

I will first show, using converging cross-modular data from allomorphy on the one hand and agreement restrictions on the other, that the suffixes to the verb base in Turoyo appear in the opposite linear order from what is expected given their underlying syntactic hierarchy: the innermost suffix is the syntactically highest, and the outermost is the lowest. (This is reminiscent of the findings of Speas 1991 and Rice 2000 for Athapaskan languages.)

Once I have motivated a basic structure for the verb word in Turoyo, I turn to a case of counterbleeding in the verbal complex. The past tense morpheme, *-wa*, has a variable linear position in the verbal complex, and interacts very differently with phonological processes as compared to morphological processes; more specifically, *-wa* is visible in its surface position to phonology, but is invisible in this position to allomorphy. I propose that *-wa* is an infix, and show how this analysis—coupled with a cyclic derivation that separates morphology from phonology—straightforwardly accounts for *-wa*’s puzzling behavior, along with the other morphological quirks of Turoyo verbs.

The analysis I put forward has several consequences for the architecture of the grammar, in particular, in arguing against incorporating allomorph choice and infixation into the phonology (contra, e.g., McCarthy & Prince 1993, Kager 1996, Wolf 2008), and in supporting cyclic, serial, realizational models that both separate syntax from exponent choice, and separate exponent choice from phonology (along the lines of Paster 2006, Embick 2010, and Bye & Svenonius 2012).

We are happy to announce that Laura Kalin will be visiting the department this week and will teach two mini courses (details below). Laura has asked that attendees read section 4 of McCarthy & Prince 1993 Generalized alignment (with the rest being optional) and chapter 3 up until the end of section 3.2 of Yu 2007 A Natural History of Infixation (the rest being optional) in preparation for the course.

Speaker: Laura Kalin (Princeton University)
Title: Theoretical approaches to infixation
Time: Wednesday, 1:00-2:30 and Thursday, 12:30-2:00
Location: 32-D461

Abstract: This mini-course will discuss a number of different theoretical approaches to infixation, and bring data from a variety of languages to bear on these approaches. More specifically, we will differentiate approaches to infixation along the following two dimensions: 
(i) Is morphology (in particular, exponent choice/suppletive allomorphy) evaluated alongside phonology, or does morphology precede phonology?
(ii) Does an infix linearly concatenate with the stem it combines with (i.e., as following or preceding the stem) before taking its surface (infixed) position inside the stem, or does an infix slot directly into its infixed position without a preliminary step of linear concatenation? 
Crossing these two dimensions gives us four logically-possible types of approaches to infixation, three of which are well-attested in the literature. On the one hand are accounts that take morphology to be simultaneous with phonology, all of which also lack a step of linear concatenation prior to infixation (e.g., McCarthy & Prince 1993, Hyman & Inkelas 1997, Wolf 2008, Samuels 2009), though these accounts do differ with respect to whether/how infixes are specified as (wanting to be) prefixes/suffixes. On the other hand are accounts that separate morphology from phonology; some such accounts take infixes to slot directly into their surface position (e.g., Yu 2007), while others include a step of linear concatenation prior to infixation (e.g., Bye & Svenonius 2012, Embick 2010). 
Using novel typological data that catalogues interactions between allomorphy and infixation, I will argue that only this last approach to infixation makes the right empirical predictions, and further, that this data supports a cyclic, serial, realizational morphological grammar.

The 3rd Crete Summer School of Linguistics will be taking place from July 13 to July 26, 2019, at the University of Crete in beautiful Rethymnon.

Current MIT faculty Adam Albright, Kai von Fintel, and Sabine Iatridou will be teaching classes at CreteLing, along with alumni Heidi Harley, Paul Kiparsky, Pritty Patel-Grosz, Douglas Pulleyblank, Philippe Schlenker, William Snyder, and Zoltán Szabó , as well as colleagues from around the world.

Full information (including details on the student application due April 6th), can be found on the school website (https://www.phl.uoc.gr/cssl19/).

Speaker:  Ljiljana Progovac (Wayne State University)
Title: Syntax in the shadow of evolution: Stepping stones into hierarchy, Move, recursion, and coordination
Time: Tuesday, Feb. 26, 1-2pm
Location: 32-D461

Abstract:
It is often stated that the nature of human language cannot be fully understood without reference to its evolution. Here I consider how one such (theoretically grounded) reconstruction of syntax may lead to a deeper understanding of attested syntactic phenomena, both those well-supported in syntactic theory (e.g. the small clause foundation of sentences), and those which remain largely unresolved (e.g. islandhood). The reconstruction leads to a proto-syntax characterized as flat, symmetric, intransitive, and static (without Move). Arguably, these rigid but robust beginnings still pervade the fabric of modern syntax, intertwined and interspersed with more modern forms, producing some intricate effects and echoes of the past, including islandhood. I will consider a variety of approximations of such proto-syntactic structures in modern languages, as well as intermediate/ambivalent forms, with the focus on the parataxis-coordination-subordination dimension (the intransitive-middle-transitive dimension was discussed in last semester’s Syntax Square talk). Both of these continuums/progressions are of relevance to the gradual emergence of the hallmarks of modern syntax, and each is characterized by ambivalent and overlapping intermediate forms. The postulated approximations of proto-syntax not only exhibit different syntactic behaviors relative to their (more) hierarchical counterparts, but they also rely on different processing strategies when investigated in neurolinguistic experiments. 
This is a practice talk for one of the three lectures to be given in Tokyo and Kyoto, Spring School in EvoLinguistics, March 2019.

Speakers: Maša Močnik & Rafael Abramovitz (MIT)
Title: Building Attitudes in Koryak
Time: Wednesday, February 27th, 1-2 pm
Location: 32-D461

Abstract: In this talk we explore the attitude verb ivyk from Koryak, a highly endangered Chukotko-Kamchatkan language of the Russian Far East. Ivyk looks like a variable-force variable-flavour attitude verb: it can be translated by (at least) say, tell/order, think, allow for the possibility, hope, fear, and wish (though not, for example, imagine). We will focus primarily on the division of labour between the doxastic readings (think, allow for the possibility) and the desiderative readings (hope, fear, wish) and argue, following Bogal-Allbritten (2016), against a lexical ambiguity or an underspecification analysis of the various readings.

Speaker: Kevin Ryan (Harvard University)
Title: Superheavy avoidance in meter
Time: Wednesday (2/27), 5:00pm-6:30pm
Location: 32-D831

Abstract:

Superheavies are known to be marked, as demonstrated by processes such as closed syllable shortening. Additionally, metrists have reported that poets disprefer locating superheavies in cadences, that is, in line endings, where the meter is strictest (e.g. Hoenigswald 1989 on Vedic and 1991 on Homeric). I verify this finding using broader and better controlled statistical tests and extend it to certain other languages. I add that superheavy avoidance, though strongest in the cadence, is evident throughout the line. Moreover, it is gradient; for instance, ultraheavies are even more strongly avoided relative to their baseline incidence.

To analyze superheavy avoidance, one could simply index *Superheavy to the relevant constituents (e.g. *Superheavy_cadence) and call it a day, as I have previously. This is unsatisfactory in a few ways. First, as such, it doesn’t capture gradient weight, though it could easily be augmented by stringent or scalar mapping to fix that (cf. Ryan 2011). Second, in general, marked structure is not avoided in cadences. For example, in Vedic, codas, retroflexes, and vowel hiatus are all marked, but none is avoided in cadences. Thus, we shouldn’t open the door to constraints like NoCoda_cadence and Onset_cadence; rather, we should explain why specifically overweight is a problem. Third, not all quantitative traditions exhibit superheavy avoidance, so we should explain its confinement to certain languages/meters. I attempt to address all of these questions through an analysis in terms of phonetic lapse (Stanton 2019).

Congratulations to fourth-year student Suzana Fong, on the publication in Glossa of her article entitled “Proper movement through Spec-CP: An argument from hyperraising in Mongolian”! Glossa is an open-access journal, so you can read the abstract and download the paper at https://www.glossa-journal.org/articles/10.5334/gjgl.667/.  

Speaker: Colin Davis (MIT Linguistics)
Title: Why extraposition is rightward
Time: Thursday, 2/28, 12:30-1:50pm
Location: 32-D461

Abstract:

In this work in progress, I’ll offer an approach to some word order puzzles about extraposition. Fox & Nissenbaum (1999) proposed that extraposition is derived by covert movement to the right, followed by late merge (Lebeaux 1999, a.o.) to that covert rightward position. Even if we accept that string-vacuous movement can in some sense have linear direction, this account does not straightforwardly apply to extraposition from phrases that undergo overt leftward movement. One of my central concerns will be this issue. I’ll argue that we can derive the possibility of extraposition from overtly moved phrases as well as covertly moved ones, without stipulating a linear direction to covert movement, through the interaction of the following concepts. #1: Spellout linearizes entire phases at once, and must not generate ordering contradictions (Fox & Pesetsky 2005, a.o.). #2: Late merge can only apply to the linear edge of a phase (NIssenbaum 2000). #3. Late merge can apply even to embedded phases (Fox 2017). Removing the hypothesis that covert movement is ordered provides insight into why extraposition fails when its source DP is deleted by ellipsis (Takahashi & Ohtaka 2017). This account leads to difficulty in deriving William’s Generalization (=the height of an extraposed phrase determines the scope of its “source”), without a more constrained notion of how extraposition is ordered, which I won’t have time to tackle here.

Speaker: Julien Musolino (Rutgers University)
Title: Studying Language Acquisition during the Preschool years: Do we need a new paradigm?
Time: Friday, 3/1, 2-3pm
Location: 32-D831

Abstract:

The preschool years have been a particularly fruitful developmental period for research on language acquisition for two main reasons. The first is that preschoolers are sophisticated language users and can therefore be tested on a range of complex linguistic phenomena of interest to linguists and psycholinguists. The second is that in spite of their linguistic savvy, preschoolers have often been found to differ systematically from adults in their linguistic behavior. Over the past several decades, these observations have led to the emergence of a dominant framework to study language acquisition during the preschool period, accompanied by a wealth of empirical findings. Its historical significance and usefulness notwithstanding, there are strong signs today that this framework needs to be fundamentally revised. In this presentation, I will introduce the standard framework, showcase some of its main applications, and discuss its limitations and the need for a new paradigm.

NYU hosted the 4th edition of the Tu+ workshop on February 16-17. Two current MIT graduate students presented talks:

• Colin Davis (4th year) on Mismatched suppletion in Azeri as morphology/phonology competition
• Ömer Demirok (5th year) on
  Ergative as a default case: Evidence from a Georgian dialect

Speaker: Stan Zompí
Title: discussion on Daniel Harbour’s 2016 book Impossible Persons
Time: Tuesday 02/19, 5-6:30pm
Location: 32-D831

In this week’s meeting of Phonology Circle, we will discuss Zuraw & Hayes’s 2017 paper: “Intersecting constraint families: an argument for Harmonic Grammar”. Erin Olson will lead the discussion. The details follow:

Discussion leader: Erin Olson (MIT)
Paper: Zuraw, K., & Hayes, B. (2017). Intersecting constraint families: an argument for Harmonic Grammar. Language, 93(3), 497-548 (available here).
Time: Wednesday 02/20, 5:00pm-6:30pm
Location: 32-D831

Abstract: 

In the analysis of free variation in phonology, we often encounter the effects of intersecting constraint families: there are two independent families of constraints, each of which has a quantifiable effect on the outcome. A challenge for theories is to account for the patterns that emerge from such intersection. We address three cases: Tagalog nasal substitution, French liaison/élision, and Hungarian vowel harmony, using corpus data. We analyze the data patterns created by intersecting families using several formal frameworks, and find that an accurate account is best based on Harmonic Grammar (in one of its two primary quantitative implementations). Our work also suggests that that certain lexical distinctions treated as discrete by classical phonological theory (e.g., “h aspiré” vs. ordinary vowel-initial words of French) are in fact gradient and require quantitative treatment.

Speaker: Beste Kamali (Universität Bielefeld)
Time: Thursday 02/21, 12:30-1:50pm
Location: 32-D461

Abstract:
Evidential bias found in certain polar questions is not well understood. I delineate a structural account based on parallel propositional alternatives, propositional alternatives to p that do not include not-p, but rather q, r etc. The evidence will come from Turkish, where evidentially biased PQs are differentiated from non-biased ones by the locus of clitic placement, and there are clearly testable intuitions of parallel propositional alternatives whenever the evidentially biased placement is used.

Speaker: Sherry Yong Chen (MIT Linguistics)
Title: The linguistics of one, two, three
Time: Thursday 02/21, 5-6pm
Location: 46-5165

Abstract:
The interpretation of number words in context is of interest not only to linguists, but also logicians, psychologists, and computer scientists. In this talk, we will discuss some theoretical and developmental questions related to the linguistics of English numerals.

Bare numerals (e.g. two, three) present an interesting puzzle to semantic and pragmatic theories, as they seem to vary between several different interpretations: ‘at least n’, ‘exactly n’, and sometimes even ‘at most n’. We will examine how the availability of a particular interpretation seems to depend on the interaction between linguistic structure and contextual factors, and discuss three approaches that try to capture the relationship between these interpretations. 

Turning to the acquisition of bare numerals, developmental research suggests that preschoolers by the age of 5 are able to access ’non-exactly’ interpretations of a bare numeral in contexts where these interpretations are licensed, just like adult speakers. A natural hypothesis for this is that the knowledge of the full range of interpretations may come through a prior understanding of the meaning of explicit expressions such as ‘at least/at most’ in English. This turns out to be questionable, however, since it is also shown that 5-year-olds haven’t yet acquired the meaning of the expressions at least and at most yet. Time permitting, we will end with a discussion about what all this means for the development of numerical concept and/or language development in general.

Michel DeGraff wrote the foreword to a recently published volume edited by Donaldo Macedo Decolonizing Foreign Language Education: The Misteaching of English and Other Colonial Languages. The foreword is entitled “Against apartheid in education and in linguistics: The case of Haitian Creole in neo-colonial Haiti “

More information about this volume can be found on Amazon and Google Play.

The materials and video recordings of the course “Creole languages and Carribean identities”, taught by Michel DeGraff in spring 2017, have been made available publicly at MIT OpenCourseWare. You can find them here.

Speaker: Colin Davis (MIT)
Title: Davis (2019) vs. Bošković (2018)
Date and Time: Tuesday, Feb. 12, 1-2pm
Location: 32-D461

In this presentation I’ll compare a paper of mine about stranding with another by Bošković about movement from moved elements, which has some overlap. A reviewer wants me to find some advantages, but it’s a bit complicated. I have some thoughts, but I hope that the audience will help. This will be an informal presentation with interruption encouraged.

Speaker: Danfeng Wu (MIT)
Title: Syntax of either in either…or… sentences
Date and time: Thursday, 2/14, 12:30-1:50pm
Location: 32-D461

Either in either…or… sentences can appear as the sister of a disjunction phrase (1), higher than the sister of a disjunction phrase (2)-(4), or lower (5)-(6), assuming the disjunction phrase is rice or beans in all these examples (based on observations by Larson (1985), Schwarz (1999), Han and Romero (2004), den Dikken (2006), a.o.).

(1) John will eat either rice or beans.

(2) John will either eat rice or beans.

(3) John either will eat rice or beans.

(4) Either John will eat rice or beans.

(5) John will either eat rice or he will eat beans.

(6) John either will eat rice or he will eat beans.

I propose an analysis of either that accounts for its distribution in these three types of sentences. I argue that there are two copies of either in an either … or … sentence. Either originates inside the disjunction phrase, and later moves (overtly or covertly) to be the sister of the disjunction phrase. In the meantime, ellipsis may delete material in the second disjunct that is identical to its counterpart in the first disjunct.

These independent ingredients of the proposal, namely the movement of either and ellipsis, may interact with each other and create complicated empirical results. In particular, I will discuss four empirical generalizations about either that the current proposal account for successfully while alternative proposals fall short.

I speculate, based on similarities between this analysis of either and previous analyses of focus-sensitive operators (e.g. Cable (2007) and Hirsch (2017)), that all focus-sensitive operators have what I call bipartite syntax: there are two instances of the operator in a sentence, one structurally higher than the other. The lower copy is semantically inert, and must c-command the focus from a local position. The higher copy agrees with a probe and/or marks semantic scope.

CompLang will resume this Spring semester. The group will be meeting on Thursdays at 5pm, in MIT room 32-370. There is something new this semester: in addition to the regular invited talks, the group will be hosting discussions on foundational topics in the language sciences. Each of these new meetings is centered around a reading, selected as a springboard for fruitful interdisciplinary discussion and debate. You can find the talk and reading schedule for this semester online on the CompLang website.

This week, the group will read and discuss “Structures, not strings: linguistics as part of the cognitive sciences”. Details follow below. 

Paper to read: Structures, not strings: linguistics as part of the cognitive sciences by Everaert et al. (2015)

Presenter: Peng Qian (MIT BCS)

Date and time: Thursday, 2/14, 5-6pm

Location: 32-370

Details: This review paper views human language as a computational system in the mind, primarily for the expression of thoughts. With several examples from semantics, syntax, morphology, and phonology, this paper illustrates the central role of hierarchical structures, not linear properties of strings, in formulating an explanatory and theoretical characterization of linguistic knowledge. 
We would like to invite you to join us in thinking and discussing the insights and critical aspects of the arguments outlined in this review paper, and how different perspectives, alternative approaches, and interdisciplinary methodologies could enrich our inquiry and understanding of human language.

The East Coast Syntax Workshop (ECO5) took place at the University of Maryland this year on Saturday (February 9th). Suzana Fong (4th year) was there to represent MIT ; she gave a a talk entitled “CONCORD vs. INDEX number in Wolof bare nominals ”. She reports: “Not only did I get useful feedback on research in progress, I was also happy to see again fellow linguists from past editions of the workshop and meet UMD students and faculty.”

Welcome to the first edition of Whamit! for Spring 2019! 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, Keny Chatain, Tracy Kelley, Elise Newman, HyunJi Yoo.

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

Best wishes for the new year!

Visiting Scholar

Chen Zhao (Huazhong University of Science and Technology)

I’m currently teaching French and introductory courses of linguistics to undergraduates majored in French at Huazhong University of Science and Technology (China). My research is grounded in formal syntax, with a special interest in labeling theory, Linearization, object shift, Syntax of Chinese and typological studies of different languages under the generative framework. My favorite part of the research is to be able to shed new lights on a linguistic phenomenon that has already been thoroughly studied by the scholars. I am currently involved in an investigation of the symmetry in the syntax and different strategies that grammars resort to in order to label the symmetric structures.

Visiting Student

Gregor Williamson (University College London)

I’m from London, UK. I did an undergraduate degree in English Language teaching. I did an MA in theoretical linguistics at UCL, where I am currently doing a PhD. The majority of my research is concerned with the syntax-semantics of the clausal spine (VP, TP, CP). I am particularly interested in various types of clausal embedding (non-finite clauses, adverbial clauses, attitude reports). My interests outside of linguistics include: playing the piano and climbing.

Speaker: Geoffrey Schwarz (Adam Mickiewicz University)

Title: There is no such thing as [voice] - evidence from Polish and Polish-accented English

Date/Time: Wednesday (2/6), 5:00pm-6:30pm

Location: 32-D831

Abstract: available here

Speaker: laksdjl
Title:
Time: , 5pm - 6:30pm
Location: 32-D831

Abstract: joshed

Speaker: ftjftuf
Title: utffut
Time: , 5pm - 6:30pm
Location: 32-D831

Abstract: ftutut

• 24.942: Topics in the Grammar of a Less Familiar Language
• 24.947: Language Disorders in Children
• 24.954: Pragmatics in Linguistic Theory
• 24.960: Syntactic Models
• 24.964: Topics in Phonology
• 24.966J: Laboratory on the Physiology, Acoustics, and Perception of Speech
• 24.979: Topics in Semantics
• 24.981: Topics in Computational Phonology

24.942 Topics in the Grammar of a Less Familiar Language
Kenstowicz, Richards
Students will work with a native speaker of Javanese, examining aspects of its syntax, semantics, and phonology. In the course of doing this, students will acquire techniques for gathering linguistic data from native speakers.

24.947 Language Disorders in Children
Flynn
Reading and discussion of current linguistic theory, first language acquisition and language disorders in young children. Focus on development of a principled understanding of language disorders at the phonological, morphological and syntactic levels. Examines ways in which these disorders confront theories of language and acquisition.

24.954 Pragmatics in Linguistic Theory
Fox, Levy
Formal theories of context-dependency, presupposition, implicature, context-change, focus and topic. Special emphasis on the division of labor between semantics and pragmatics. Applications to the analysis of quantification, definiteness, presupposition projection, conditionals and modality, anaphora, questions and answers.

24.960 Syntactic Models
Pesetsky
Comparison of different proposed architectures for the syntax module of grammar. Subject traces several themes across a wide variety of approaches, with emphasis on testable differences among models. Models discussed include ancient and medieval proposals, structuralism, early generative grammar, generative semantics, government-binding theory/minimalism, LFG, HPSG, TAG, functionalist perspectives and others.

24.964 Topics in Phonology
Steriade
Description: The goal of this class is to understand some of the results of linguistic research using the Artificial Grammar paradigm (Reber 1967, 1993).  The AG studies we consider compare learning outcomes between subject groups exposed to artificial languages that are designed to differ just in whether they present ‘natural’ or ‘unnatural’ patterns. The definition of ‘naturalness’ varies, and understanding what should be meant by this term will be one focus of discussion. Naturalness is commonly defined in terms of attestation – with natural taken to mean well-attested – but this needs rethinking. 
Most of the large body of phonological AG work now available dates from the last 15 years. The hypothesis commonly tested is that the naturalness of the target pattern is a determinant of learning success. When this hypothesis is supported, subjects succeed in learning the natural pattern and fail with the unnatural one. In many studies, however, null results are obtained: subjects acquire to the same extent patterns identified as natural and unnatural. We will begin the course by exploring one interpretation of these mixed results, due to Moreton and Pater 2012a,b, and Moreton 2008: the successfully learned natural patterns are not necessarily natural in a typological, or phonetic sense, but they are simpler than the patterns that the subjects fail to learn. This idea comes with its own learning algorithm and with a distinct view of how phonological typology relates to competence, or doesn’t. It is a useful way to explore large parts of the AG literature but we will see that it has limitations. A different idea to consider is that studies that have successfully shown that natural patterns are easier to learn than unnatural ones have focused on faithfulness, or can be reinterpreted in those terms. The null results are found in the markedness domain.  If this interpretation continues to look promising, we’ll look for an explanation.
There are other potentially interesting results in the phonological AG work and we may expand the reading list and consider syntax if participants express an interest. I can’t promise answers to any of the questions raised above, but the best of this AG body of literature tackles two essential questions in the field: do aspects of a speaker’s grammatical knowledge reflect the linguistic typology? what are the sources of this knowledge?

24.966J Laboratory on the Physiology, Acoustics, and Perception of Speech
Braida, Shattuck-Hufnagel, Choi
Experimental investigations of speech processes. Topics include computer-aided waveform analysis and spectral analysis of speech; synthesis of speech; perception and discrimination of speech-like sounds; speech prosody; models of speech recognition; speech development; analysis of atypical speech; and others. Recommended prerequisite: 6.002, 18.03, or 24.900.

24.979 Topics in Semantics
Fox, Hackl, Schwarzschild
The seminar will focus on topics in degree semantics. Issues to be addressed include:

• how degree operators interact with individual and modal quantifiers 
• interpretation of degree modifiers of noun phrases (e.g. at least)
• analysis of relative clauses modifying nominalized degree predicates (height, weight)
• whether degrees are primitive or derived 
• analysis of degree constructions using vectors or directed segments
• external and internal syntax of comparative clauses

Our starting point will be Schwarzschild 2019, which we will later try to use as the basis for the discussion of older literature. Some of the sessions will be lead by registered students in collaboration with the instructors. 

24.981 Topics in Computational Phonology
Albright
Computational modeling can usefully inform many aspects of phonological theory. Implementing a theory provides a more rigorous test of its applicability to different data sets, and requires a greater degree of formal precision than is found in purely expository presentations. By training learning models on realistic training samples, we can test whether a posited analysis can actually be discovered from representative data, and we can observe what proportion of the data is actually accounted for by that analysis. Modeling also provides a direct means of testing whether a proposed formal device facilitates the discovery of generalizations, or whether it hampers learning by greatly increasing the size of the search space. In the most interesting cases, computational modeling uncovers facts about the language that would have been difficult to discover by eye, and forces us to ask which facts are treated as linguistically significant by speakers.
Topics will include: (subject to revision)
- Statistical “baseline” models (n-gram models, exemplar models)
- Algorithms for constraint ranking and weighting
- Algorithms for constraint discovery
- Integrating learned and innate constraints
- Learning in the midst of variation and exceptions, and discovery of gradient patterns
Requirements: readings and small regular problem sets, final project+presentation.

Speaker: Martina Wiltschko

Title: How to do things with nominals. Towards a syntax of nominal speech acts
Time: Friday, February 8th, 4pm-5:30pm

Room: 32-155

Abstract: Going back to Aristotle, classic grammatical description as well as current theories of grammar and the construction of meaning take the sentence to be the object of investigation. In his seminal work, Austin 1962 took a first step towards breaking with this tradition within philosophy of language. He argued that our understanding of meaning has to be informed by the fact that when we say things, we also do things. Different types of sentences give rise to different speech acts such as asserting, questioning, requesting, promising etc. and over the past sixty years, evidence has accumulated suggesting that speech act meaning is part of sentences, and hence part of grammatical structure. However, when we do things with words, we do not always use sentences. For example, a content question may be answered with a nominal phrase only (e.g., Who wrote that essay? Penelope). And by uttering a nominal we are also doing something. Specifically, we are attempting to identify a referent in a manner that will enable our interlocutors to recognize them.

This leads us to postulate a novel hypothesis, namely that noun phrases (like sentences) may be full-fledged speech acts. We refer to this as the nominal speech act hypothesis. Specifically, we postulate that, just like clauses, nominals are dominated by a layer of structure which encodes pragmatic information contributing to their use-conditions. They differ in that the speech act layer in the clausal domain establishes a direct link between the speech-act participants and the proposition denoted by the clause, while the speech act layer in the nominal domain establishes a direct link between the speech-act participants and the referent denoted by the nominal.

In this talk I provide empirical evidence for a dedicated nominal speech act layer. Specifically, the evidence I discuss comes from i) cross-linguistic variation in pronominal paradigms; ii) properties of impersonal pronouns; iii) properties of formality distinctions in nominal expressions; iv) distinctions between spatial and discourse functions of demonstratives.

The Linguistic Society of America’s Annual Meeting for 2019 was held at in New York in January. As per usual, MIT was well represented. The following department members presented talks and posters:

• Tanya Bondarenko  and Colin Davis –
  Parasitic gaps diagnose concealed pied-piping in Russian
• Shumian Ye (visitor, Peking University) –
  Biased A-not-A questions in Mandarin Chinese
• Carolyn Spadine and Gunnar Lund –
  Complementizers in matrix contexts: Reporting attitudes without attitude verbs
• Juliet Stanton (alumna) and Donca Steriade (faculty) –
  Modified cyclicity: the stress of English Latinate derivatives
• Edward Flemming –
  Implications of [i] vowels for the theory of vowel inventories
• Colin Davis –
  English Possessor Extraction
• Rafael Abramovitz and Adam Albright (faculty) – 
  Morphological feature deletion without metasyncretism
• Kenyon Branan (alumnus) and Colin Davis – 
  Agreement and Unlocking at the Edge
• Veronica Boyce (BCS), Titus von der Malsburg, Till Poppels, Roger Levy (BCS) -
  Female gender is consistently under-expressed in pronoun production and under-inferred in comprehension

Alumni who presented or organised symposia include: Ezra Keshat, Michael Yoshitaka Erlewine, Jon Nissenbaum, Aron Hirsch, Michelle Yuan, Hadas Kotek, Coppe Van Urk.

This is happening this Thursday 7th from 5pm to 6pm. Martina will talk about methodologies and how to elicit interactional language (discourse markers and beyond).

Speaker: Tanya Bondarenko, Stanislao Zompi’
Title: Leftover Agreement: spelling out number in Kartvelian languages
Date and time: Monday 12/10, 5-6pm
Location: 32-D831
Abstract:

In this talk, we present our work in progress in which we argue that sometimes probes agree with features on lower probes that have not yet been spelled out —- a mechanism we refer to as Leftover Agreement. We show that this mechanism is helpful in accounting for the verbal agreement paradigms in all four of the Kartvelian languages (Georgian, Laz, Megrelian, Svan). The main body of evidence comes from number agreement: we show configurations where a higher probe spells out plural features only if those features have failed to be spelled out by a lower probe. Here is a short illustration from Georgian:

(1) a.  gv-naxa                      b. g-naxa-t
          1PL-see.AOR.3SG            2-see.AOR.3SG-PL
          ‘He/she saw us.’              ‘He/she saw you (pl).’

Presenter:  Lee Bickmore (University at Albany) & Winfred Mkochi (University of Malawi)
Title:Segmental and Tonal Absolute Neutralization in CiTonga
Date/Time: Wednesday, December 12, 5:00-6:30pm
Location: 32-D831
Abstract:

CiTonga, an under-described Malawian Bantu language, exhibits a complex array of stem tone patterns. We show that these patterns can be accounted for by assuming that in addition to various lexical High tones present underlyingly, melodic (grammatical) High tones are also added. These target the stem-initial TBU if it is free, otherwise the final one. Two productive tonal processes are key to understanding the surface tones: Tone Doubling, which spreads an underlying H to the following mora, and Phrase-final Left Shift, which shifts a H off a phrase-final mora to the phrase-penultimate one. These two processes create a variety of tonal configurations, many of which violate the Obligatory Contour Principle (OCP). We show that in some cases these violations simply remain, while in other cases they are repaired. Of interest is the fact that four different repair strategies are used depending on 1) whether one or both Hs are multiply linked and 2) the domain in which the OCP violation is found (e.g. within the stem, across the stem, or across words). Finally, we present two TAMs, the Subjunctive and Imperative, which seem to have anomalous surface tone patterns, given the productive rules which have been motivated to that point. To account for these we present and discuss an abstract analysis which involves an absolute neutralization. For the subjunctive, we propose that even though the Subject Prefixes uniformly surface as Low, they must be set up as High in order to trigger the various processes which directly account for the surface stem tone patterns, even though this H is not ultimately realized. For the imperative, a segmental prefix is posited in order to trigger the appropriate processes, which must ultimately be deleted (i.e. both the mora and tone of the prefix). We conclude by briefly comparing this analysis to several alternatives.

Whamit! will be on its Winter (semi-)hiatus from now until the start of the Spring semester. Weekly posts will resume on February 5th, 2019. In the mean time, we will have rolling posts, publishing breaking MIT Linguistics news as it happens. Thanks to all our contributors, editors, and you dear readers!

See you next year!

We are delighted to announce that Emily Elfner will be here for an extended visit, during which she will give a mini course in two parts, details below.

Speaker: Emily Elfner (York University)
Time: Wednesday, December 5th, 1pm-2:30pm; Thursday, December 6th, 12:30pm-2pm
Place: 32-D461
Abstract:

In recent years, there has been considerable renewed interest in debates regarding the syntax-prosody interface and the source of prosodic domains. Broadly speaking, there are two types of approaches to the “discovery” of prosodic domains: an “intonation first” model, which derives prosodic domains on the basis of observable phonological processes (such as intonational patterns, gradient phonetic markers to prosodic boundaries, and domain-sensitive phonological processes), and a “syntax first” model, which derives prosodic domains on the basis of syntactic constituent structure. Within these two extremes, most approaches to the syntax-prosody interface will assume a middle ground, with an implicit assumption that both syntax and phonology should be taken into account in the positing and diagnosing of prosodic domains. However, modelling the exact contribution of each part has proven to be difficult to determine and subject to some debate.

In this mini-course, I will provide an overview and critical analysis of a current “hybrid” approach to the syntax-prosody interface, Match Theory (Selkirk 2011), which incorporates a direct approach to syntax-prosody mapping under the guise of Prosodic Hierarchy Theory. Match Theory may be considered to be a “syntax first” model because it assumes that the origin of prosodic domains lies in the syntactic component, and that these are mapped onto the phonological component, specifically, the prosodic hierarchy, via a family of violable MATCH constraints. However, because these constraints are violable and evaluated in the phonological component (as in an OT model), mismatches, as observed via domain-sensitive phonological processes, can be accounted for systematically using violable prosodic markedness constraints that interact with the relevant MATCH constraints. Thus, while prosodic domains are mapped directly from syntactic structure, resulting in recursive prosodic domains, phonological and intonational processes can still be used as evidence to help diagnose and define the edges and scope of prosodic domains. The positing of prosodic domains is thus approached from both sides (syntax and phonology), meeting in the middle. This approach to the syntax-prosody interface has important consequences for our understanding of the syntax-prosody interface and the continuing role of the prosodic hierarchy in accounting for phrase-level phonological processes.

SNEWS happened this Sunday at UMass Amherst.

• Christopher Baron (3rd year) presented “Different measures for degree achievements”. His hand-out can be found here.
• Filipe Kobayashi and Sherry Chen (2nd year) presented “Are adnominal and adverbial distributive numerals the same? Perspectives from Mandarin Chinese.” Their handout can be found here: https://sherrylinguist.files.wordpress.com/2018/12/numnum_snews2018.pdf

Speaker: Emma Nguyen (UConn Linguistics)
Title: Using developmental modeling to specify learning and representation of the passive in English children
Date and time: Thursday, 12/6, 5-6pm
Location: 46-5165
Abstract:

Complete knowledge of the passive takes significant time to develop in English children. Building on prior work identifying the importance of lexical factors for this process, we specify a Bayesian learning model that can capture experimentally-observed passivization behavior in five-year-olds, given child-directed speech to learn from. Through this developmental model, we identify (i) how English children may be integrating lexical feature information, and (ii) how costly they may view the passive structure to be.

The next meeting of the fieldwork group is happening on Thursday. Emily Elfner will talk about recording procedures and equipment.

Time: Thursday, 12/6, 5-6pm
Place: 32D-831
Speaker: Emily Elfner
Topic: Recording in the field

Speaker: Emily Elfner (York University)
Title:  The Syntax-Prosody Interface
Time: Friday, December 7th, 3:30-5pm
Place: 32-155
Abstract:

In this talk, I discuss word- and phrase-level prominence in two languages which may be characterized as “polysynthetic”: Kwak’wala (North Wakashan: BC, Canada) and Inuktitut (Eskimo-Aleut; Northern Canada). Polysynthetic languages are typically described as such in terms of their morphosyntactic characteristics: allowing complex morphological “words” which blur the lines between words, phrases, and clauses. In recent years, researchers have argued that complex words may be analyzed using the same syntactic tools used for less agglutinative languages, proposing that the unusually complex “words” result from the ways in which such structures are spelled out phonologically (e.g. Compton & Pittman 2010). This assumption places the onus on the syntax-phonology interface to account for notions like what constitutes a word, notions that were previously thought to be determined by a language’s morphological and syntactic properties.

Kwak’wala and Inuktitut share similarities in terms of their word-level systems of prominence: both show evidence of a single (tonal) prominence within each word-level domain, but little evidence of prosodic structure within that domain: no secondary stress or evidence of alternating metrical structure, as would be expected of foot structure, nor any apparent sign of word-internal morphophonological domains. What does this mean for theories of prosodic structure in which prosodic constituents are derived from syntactic structure? In other words, why do “words” spell-out as large domains in languages like Inuktitut or Kwak’wala, but smaller domains in languages like English? This talk will explore some preliminary answers to these questions on the basis of these two case studies.

Partly based on joint work with Patricia A. Shaw (Kwak’wala), Anja Arnhold (Inuktitut), and Richard Compton (Inuktitut).

Speaker: Rafael Abramovitz
Title: Successive-Cyclic Wh-Movement Feeds Case Competition in Koryak
Date and time: Monday 11/26, 5-6pm
Location: 32-D831
Abstract:

Recent debate surrounding theories of ergative case has centered on two types of analyses: ergative as a dependent (configurational) case (Yip et. al. 1987, Marantz 1991, Baker 2015, a.o.), and ergative as an inherent case (Nash 1996, Woolford 1997, a.o.). On the former, ergative case is assigned to the external argument of a transitive verb by case competition: it `competes’ for case assignment with another nominal in the same phase, and is assigned ergative because it is the higher of the two. On the latter, ergative is assigned to the external argument of a transitive verb by being merged as the specifier of an agentive vP. In this paper, I present new evidence for the configurational analysis of ergative case from Koryak (Chukotko-Kamchatkan), showing that movement of absolutive wh-words into higher case domains triggers dependent case in the higher domain, as well as in intermediate domains along the movement path. This is easily accounted for on the dependent case analysis, but cannot be captured if ergative marking is tied to thematic roles, as the inherent case view holds.

We are delighted to announce that Martina Martinovic will be here for an extended visit, during which she will give a mini course in two parts, details below.

Speaker: Martina Martinovic (University of Florida)
Time: Wednesday, November 28th, 1pm-2:30pm; Thursday, November 29th, 12:30pm-2pm
Place: 32-D461
Abstract:

It is a long-standing observation that syntactic features can be differently distributed over the functional spine. The strong cartographic approach commonly follows the “one property, one feature, one head” principle (e.g., Kayne 2005; Cinque and Rizzi 2010), however, the reality of cross-linguistic variation has made it clear that the bundling of features often cannot be achieved via traditional syntactic processes (such as head movement). Various bundling parameters and other mechanisms have been proposed to account for the variation in feature distribution (e.g. the “Split-IP” parameter, Thráinsson 1996, Bobaljik & Thráinsson 1998; the “Voice-bundling” parameter, Pylkänen 2002, 2008; feature “scattering”, Giorgi & Pianos 1996; etc.). Particularly popular in the last decade have been modifications of the Feature Inheritance model  (Chomsky 2005, 2008; Richards 2007, 2011), proposing that Feature Inheritance is obviated under certain conditions (e.g. Ouali 2006, Fortuny 2008, Legate 2011, Gallego 2014, Aldridge 2015, 1017).

In this mini-course we will focus on the properties of the clause-peripheral layers, commonly known as CP and TP, the cross-linguistic variation in how features traditionally associated with C and T are distributed over functional heads, and how this advances our understanding of how syntactic structures are built. The bulk of the data will come from Wolof and will explore the Head-Splitting approach developed in Martinović 2015. We will discuss recent extensions of this proposal (e.g. Erlewine 2018), and other data that might be amenable to similar treatment, with the goal of identifying a possible unified approach.

Martina Martinović is going to answer all your questions about doing linguistics in the field and her personal fieldwork experience.

Time: Thursday Nov 29th, 5-6pm
Place: 32-D831
Speaker: Martina Martinović (University of Florida)

Speaker: Martina Martinovic (University of Florida)
Title: Bi-clausal progressives in Wolof (partly joint work with Marie-Luise Schwarzer)
Time: Friday, November 30th, 3:30pm-5pm
Place: 32-155
Abstract:

In many languages, clauses predicating location and progressive constructions are related. Progressives often develop from or contain locative markers/copulas, or entire locative constructions (Heine & Reh 1984, Heine et al. 1991, Demirdache & Uribe-Etxebarria 2000, Heine & Kuteva 2002). Much work has shown that progressive constructions are often bi-clausal, consisting of a locative clause that embeds a nominalized complement clause or an adjunct (Comrie 1978, Bybee et al. 1994, Fontanals & Simon 1999, Polinsky & Comrie 2002, Laka 2006, Salanova 2007, Coon 2010).

In this talk I discuss the syntax of default progressive construction in Wolof, which contain the element angi, in the literature most commonly considered to be progressive aspect. I reanalyze angi as a bimorphemic element composed of the A′-complementizer and a locative clitic. This is supported by the fact that progressive clauses can be extracted out of. Next, I propose that progressives in Wolof are biclausal, consisting of a locative clause that embeds a reduced imperfective infinitival clause. I give two pieces of evidence for this claim. First, PP modifiers can only follow the main verb in mono-clausal constructions in Wolof. In progressive constructions, however, they can precede the verb, suggesting that they are modifying a higher predicate. Second, progressive constructions cannot be negated, which would be puzzling under a mono-clausal analysis. A bi-clausal analysis straightforwardly accounts for this fact, because the two clauses that the progressives consist of — the locative clause and the reduced imperfective infinitival clause —  independently cannot contain negation. This work gives further cross-linguistic support for the bi-clausality of progressive structures, and enriches the typology of bi-clausal progressives.

Michel DeGraff was at the University of Hawaii at Hilo (@UHHilo) during the week of November 12, visiting their Program in Linguistics which is part of the College of Hawaiian Language. Michel also gave a talk there on ” Haitian Creole for education, human rights & development.” The Program in Linguistics at UH Hilo has set up an inspiring example of collaboration among linguistics, language revitalization and teacher education. The Hawaiian language was nearly exterminated through US cultural genocide starting in the late 19th century when US marines overthrew the Hawaiian monarchy. Then the US banned the use of the Hawaiian language. In the early 1980s there were fewer than 50 children speaking Hawaiian. But now Hawaiian has been revitalized and counts some 10,000 speakers, especially thanks to the efforts of the College of Hawaiian Language’s laboratory schools, which start as early as in pre-schools (called “Pūnana Leo” in Hawaiian; i.e., “language nests”). These laboratory schools may well be the best examples of language-immersion schools in the context of language revitalization. For a sampling of images of these immersion activities, both at @UHHilo and in the laboratory schools, see Michel’s videos and photos on Facebook:

• Welcoming ceremony at College of Hawaiian Language @UHHilo: https://www.facebook.com/michel.degraff/videos/10156903187358872
• Welcoming ceremony at Nāwahī School (Hawaiian language-immersion school): https://www.facebook.com/michel.degraff/videos/10156903190423872
• Hawaiian Language Class @UHHilo: https://www.facebook.com/michel.degraff/videos/10156917062308872/
• Morning Assembly at College of Hawaiian Language @UHHilo https://www.facebook.com/michel.degraff/videos/10156917418418872/
• Classes at the Nāwahī School (Hawaiian language-immersion school): https://www.facebook.com/michel.degraff/posts/10156907815508872
• Conversation with Dr. Keiki Kawai’ae’a (the Director of the College of Hawaiian Language) and her daughter (Kananinohea Māka’imoku) : https://www.facebook.com/michel.degraff/posts/10156898778408872
• Photos in and around @UHHilo: https://www.facebook.com/michel.degraff/posts/10156899735353872
• Photos at Nāwahī school: https://www.facebook.com/michel.degraff/posts/10156903514133872
• A sampling of Hawaiian music and dance: https://www.facebook.com/michel.degraff/videos/10156894663608872/

Abigail Thwaites will be presenting a paper by Torres on De Se content. The meeting will take place on Monday 19th in the 8th floor seminar room.

Title : In Defense of De Se Content

Author(s) : Stephan Torre

Abstract :

There is currently disagreement about whether the phenomenon of first-person, or de se, thought motivates a move towards special kinds of contents. Some take the conclusion that traditional propositions are unable to serve as the content of de se belief to be old news, successfully argued for in a number of influential works several decades ago. Recently, some philosophers have challenged the view that there exist uniquely de se contents, claiming that most of the philosophical community has been under the grip of an attractive but unmotivated myth. At the very least, this latter group has brought into question the arguments in favor of positing special kinds of content for de se belief; I think they have successfully shown that these arguments are not as conclusive, or fully articulated, as many have taken them to be. In this paper I will address these challenges directly and I will present and defend an argument for the conclusion that the phenomenon
of de se thought does indeed motivate the move to a special kind of content, content that is uniquely de se.

Speaker: Rafael Abramovitz
Title: Outward-Looking Phonologically Conditioned Allomorphy in the Koryak Verb: A Conspiracy in Exponence.
Date and time: Monday 11/19, 5-6pm
Location: 32-D831
Abstract:

“Bobaljik (2000) proposes that phonologically-conditioned allomorphy (PCA) can only be inward-sensitive. The reasoning goes like this: Assume vocabulary insertion is cyclic. Given this, when a given node is undergoing vocabulary insertion, phonological information will not be present for any nodes above it. This precludes the choice of the vocabulary item from being influenced by the shape of any higher morphemes. In this talk, I will discuss a counterexample to this proposal from a conspiracy in exponence found in the verb-word of Koryak (Chukotko-Kamchatkan; Russian Far East), in which an aspectual (?) suffix and an agreement suffix cannot coexist if the agreement suffix begins in a non-coronal consonant. In intransitive verbs, this causes the (outer) agreement suffix to delete, making this an instance of inward-looking PCA. In transitive verbs, however, the (inner) aspectual suffix deletes, making this an instance of outward-looking PCA. I analyze this by appealing to obliteration, an operation that deletes entire terminals, but, following Vergara & Luis (2017), argue that this obliteration takes place after vocabulary insertion. To account for the conspiracy involved in the realization of these morphemes, I propose that this case of obliteration is governed by a grammar of ranked, violable constraints (Prince & Smolensky 2008). This analysis is consistent with the predictions about the directionality of allomorphy based on cyclic spellout because it shifts the responsibility for this alternation to a post-spellout operation. Time permitting, I’ll talk about the difficulty of assigning specific semantic contributions to elements we might want to call tense and aspect morphemes in the Koryak verb to show why I put a question mark in the phrase ‘aspectual (?) suffix’ above.”

Speaker: Keny Chatain (MIT)
Title: “The same” = the + same ?
Date and time: Wednesday, November 21, 1-2 pm
Location: 32-D461

Abstract:      

This is work in progress on the semantics of same; I will need your indulgent feedback! Some authors (Barker 2007, a.o.) have puzzled at the obligatory presence of the in (1-3), whose effect on the semantics is not readily detectable, and is unnecessary given certain assumptions about the meaning of same. If this is so, then same enters a small family of operators (along with superlatives, and only), which mysteriously seem to weaken the standard Fregean presupposition of the definite.

(1) Alison praised the same poet as Carlos.
(2) Alison and Carlos praised the same poet.
(3) Every librarian praised the same poet.

In light of this problem, Charnavel (2011) notices a peculiarity of French: in a limited set of environments, French allows a same “un même”. One crucial difference that Charnavel reports between French a same and the same is a (barely recognizable) Fregean presupposition. Building off of this observation, I will: a) identify the detectable contribution the makes to (1-3) in English, using NPI data, b) show that in a scoping account of same à la Barker, this contribution is readily derived from nothing but the standard semantics of the. Finally, I will present some ideas for unifying the use of same in (1) and in (2-3).