The Weekly Newsletter of MIT Linguistics

Issue of Monday, March 14th, 2016

ESSL/LacqLab 3/14 - Katsuo Tamaoka

Speaker: Katsuo Tamaoka(Nagoya University, Japan)
Title: Indexing Movement: Eye-tracking experiments on Japanese scrambled sentences
Date/Time: Monday, March 14, 5:00 pm – 6:00 pm
Location: 32-D769 (location change)

Movement is part of many syntactic structures. However, due to the absence of a directly-compatible baseline, it is usually difficult to visualize syntactic movement in sentence. Japanese scrambled sentences provide an ideal environment for visualizing movement. This talk presents the results of eye-tracking experiments that compared scrambled and canonical Japanese sentences. The results indicate three possible indexes for movement; (1) re-reading time of a moved phrase, (2) regression frequency ratio into the moved phrase, and (3) regression frequency ratio out of the possible trace position. At least one of these three (if not all), depending on the ease of processing load, is likely to appear in the scrambled sentences. This talk will propose the possible indexes for movement in the cognitive processing of sentences.

Phonology Circle 3/14 - Patrick Jones and Jake Freyer

Speaker: Patrick Jones (Harvard) and Jake Freyer (Brandeis)
Title: Emergent complexity in melodic tone: The case of Kikamba
Date/Time: Monday, March 14, 5:00-6:00pm
Location: 32-D831

Melodic tone assignment, in which inflectional features of verbs are signaled entirely through tonal morphemes assigned to particular positions within verb stems, are pervasive within Bantu languages. A considerable body of recent work has focused on melodic tone in various Bantu languages, in an effort to better understand its core properties (in particular, the extensive 2014 volume of Africana Linguistica, edited by Lee Bickmore and David Odden). From this work, one possible conclusion is that melodic tone is relatively unconstrained both in what tones it may assign and what positions within the verb stem they may target. For example, in one extreme case, in Kikamba, a melody reportedly assigns four distinct tones to three separate positions simultaneously. In this talk, we propose a reanalysis of Kikamba which (a) restricts melodies to two target positions and (b) reduces the total inventory of target sites. More generally, we argue that since core properties of melodic tone are often obscured in surface forms due to interactions with language-particular rules, the cross-linguistic comparison of melodic tone should proceed on the basis a (more) underlying level in which these rules are controlled for.

Syntax Square 3/15 - Abdul-Razak Sulemana

Speaker: Abdul-Razak Sulemana (MIT)
Title: Deceptive Overt wh-movement in Bùlì
Date: Tuesday, March 15th
Time: 1:00pm-2:00pm
Place: 32-D461

Deceptive Overt wh-movement in Bùlì: Abstract

LFRG 3/16 - Aron Hirsch

Speaker: Aron Hirsch
Time: Wednesday, March 16, 1-2pm
Place: 32-D831
Title: Co-ordinating questions

There is disagreement in the literature about whether or not constituent questions can be disjoined (e.g. Groenendijk & Stokhof 1989, Szabolcsi 1997, Krifka 2001, Haida & Repp 2013). In this talk, I introduce novel data arguing that questions can be both disjoined and conjoined. I propose an analysis of disjoined constituent questions which unifies them with alternative questions and mention-some questions. Although questions are often taken to denote sets of propositions (type ), the proposed analysis will further render the data compatible with a hypothesis that only type t meanings can be co-ordinated (e.g. Hirsch 2015, Schein 2014, Ross 1967).

Ling Lunch 3/17 - Polina Berezovskaya

Speaker: Polina Berezovskaya (University of Tübingen)
Title: `Small Degrees’: Degree Modification in Nenets.
Time: Thursday, March 17th, 12:30-1:45 pm
Place: 32-D461

In Nenets, an underrepresented Samoyedic language from the Uralic language family, the suffix ‘-rka’ is commonly found on the gradable adjective in comparison constructions, cf. (1):

(1) Katja Masha-xad saml’ang santimetra-nh pirc’a-rka.
Katja Masha-ABL five cm-DAT tall-RKA
‘Katja is a little taller than Masha.’

Contrary to claims in the descriptive literature (e.g. Terezhenko 1947, Nikolaeva 2014), according to which this suffix is a comparative marker, original fieldwork data shows that it is not. I am proposing an analysis under which ‘-rka’ is a degree modifier that modifies a difference degree stating that this degree is small.

The Puzzle. The striking fact is that this suffix also appears outside of comparison constructions. (2) and (3) show instances of ‘-rka’ on nouns and verbs.

(2) a. ngamderc’- chair, ngamderc’arka - kind of a chair
b. neb’a - mother, neb’arka - a mother who kind of fulfils her duties as a mother, but not quite
c. ne - woman, nerka - kind of a woman (e.g. doesn’t really behave like one)

(3) Man’ s’urba-rka-dm.
I run-RKA-1.SG
‘I ran a little.’

The question here is what the core meaning contributed by ‘-rka’ in all these cases is and whether nouns and verbs carrying this suffix make reference to some kind of an implicit comparison.

The Plot. In this talk, I will explore the meaning contribution of this suffix in and outside of comparison constructions. An analysis of Nenets comparisons that are not marked by ‘-rka’ will be provided in the spirit of the standard analysis (cf. von Stechow 1984, Heim 2001, Beck 2011). I will also propose an analysis of cases like (1) using a certain type of the Restrict operation (cf. Chung and Ladusaw 2004). Ultimately, we will discuss the meaning contribution of -rka outside of comparisons.

Fieldwork Meeting 3/16 - Norvin Richards

Speaker: Norvin Richards (MIT)
Title: Fieldwork Methodology and More
Date: Wednesday, March 16th
Time: 5-6pm (UPDATED - one hour later than originally announced)
Place: 7th floor seminar room

Norvin will be talking to us about a few fieldwork-related issues (elicitation preparation, types of judgment, etc.)

A few topics that he might cover are:

(i) judgment types (grammaticality, felicity, etc.), judgment scale (good/?/*, numeric, etc.), ways to elicit judgments

(ii) elicitation prep: questionnaires with randomized items, grouping questions that target the same phenomenon together (not randomized), etc.

(iii) ways to elicit specific constructions, diagnostics to detect certain phenomenon

(iv) how to recruit speakers: any criteria for selecting speakers to work with when one just starts on a language (gender, education background, occupation, etc.).

(v) his fieldwork in Australia

3/16 - Naomi Feldman

Speaker: Naomi Feldman (University of Maryland)
Title: How phonetic learners should use their input
Time: Wednesday, 03/16/2016, 3:30-5:00 pm
Venue: 32D-461

Children have impressive statistical learning abilities. In phonetic category acquisition, for example, they are sensitive to the distributional properties of sounds in their input. However, knowing that children have statistical learning abilities is only a small part of understanding how they make use of their input during language acquisition. This work uses Bayesian models to examine three basic assumptions that go into statistical learning theories: the structure of learners’ hypothesis space, the way in which input data are sampled, and the features of the input that learners attend to. Simulations show that although a naïve view of statistical learning may not support robust phonetic category acquisition, there are several ways in which learners can potentially benefit by leveraging the rich statistical structure of their input.

Colloquium 3/18 - Naomi Feldman

Speaker: Naomi Feldman (University of Maryland)
Modeling language outside of the lab
Time: Friday, 03/18/2016, 3:30-5:00 pm
Venue: 32D-461

Speakers and listeners operate in complex linguistic environments. They extract phonetic information from highly variable speech signals and track the salience of entities in rich discourse contexts. However, little is known about the representations that support language use in these complex environments. In this talk, two cognitive models that were developed for laboratory settings are modified to operate over more naturalistic corpora. A rational speaker model is used to predict how entities are referred to in news articles, and a model of speech perception is trained and tested directly on speech recordings. In each case, simulation results show how cognitive modeling can be used to probe the way in which speakers and listeners represent the complexity of their linguistic environment.