UMass faculty are on the move this week, giving colloquia around the world!
Gaja Jarosz is giving a colloquium talk at the MIT Department of Linguistics on December 3rd at 3:30PM. Her talk is entitled ‘Learning Hidden Phonological Generalizations ‘. More information on the talk can be found here.
Kristine Yu is giving a colloquium talk at the Rutgers Department of Linguistics on December 3rd, at 3:00PM. Her talk is entitled ‘Building phonological trees.’ More information on the talk, including details on how to register for the Zoom event, can be found here.
Brian Dillon is giving a colloquium talk at the Tel Aviv University Department of Linguistics on December 2nd at 9:15AM. His talk is entitled ‘Grammatical constraints on reference: The view from comprehension.’ More information on the talk can be found here.
Abstracts are found below!
Learning Hidden Phonological Generalizations, Gaja Jarosz: Language acquisition proceeds on the basis of incomplete, ambiguous linguistic input, and one source of this ambiguity is hidden phonological structure. Due to recent developments in computational modeling of phonological learning, there now exist numerous approaches for learning of various kinds of hidden phonological structure from incomplete, unlabeled, and noisy data. These computational models make it possible to connect the full representational richness of phonological theory with noisy, ambiguous corpus data representative of language learners’ linguistic experience to make detailed and experimentally testable predictions about language learning and generalization. In this talk, I briefly review these computational developments and then discuss two ongoing projects that utilize these mutually-informing connections between computation, phonological theory, and experimental data to test hypotheses about the abstract representations that underlie phonological knowledge.
Building phonological trees, Kristine Yu: Computational perspectives from string grammars have richly informed our understanding of phonological patterns in natural language in the past decade. However, a prevailing theoretical assumption of phonologists since the 1980s has been that phonological patterns and processes are computed on trees built with prosodic constituents such as syllables, feet, and prosodic words. Moreover, multiple dependencies in prosodic structures, such as multiple association of a tone to a higher-level prosodic node in addition to a tone bearing unit such as a mora or syllable, have been broadly assumed in intonational phonology without much comment. We revisit these concepts and show that multiple bottom up tree transducers provide a natural representation for multiple tonal association as well as multiple dependencies in prosodic structures in general, including prosodically-conditioned segmental allophony.
Grammatical constraints on reference: The view from comprehension, Brian Dillon: Experimental research has consistently shown that the grammatical knowledge reflected in (e.g.) the Binding Theory guides real-time pronoun interpretation. Evidence for this conclusion comes from a range of experimental evidence that comprehenders selectively activate grammatically-licit antecedents when processing pronouns and anaphors. But what are the mechanisms by which comprehenders arrive at these grammatically constrained interpretations? And what is the relationship that these mechanisms bear to offline grammatical knowledge? Psycholinguistic studies of pronoun interpretation emphasize the fact that pronoun interpretation is a mixture of bottom-up and top-down processes (e.g. Rohde & Kehler, 2019). In this talk I aim to show that this perspective offers a useful framework for thinking about how structural constraints on coreference enter into the real-time comprehension of pronouns and anaphors. Using Principle B effects as a test case in a range of languages, I will survey a number of experimental studies done with our research group here at UMass and collaborators at the University of Toronto that illustrate how top-down and bottom-up pronoun resolution processes jointly contribute to disjoint reference phenomena in pronoun interpretation.