27th October 2022 04:30 PM

Linguistics Colloquium Speaker: David Goldstein

The Department of Linguistics proudly presents Dr. David Goldstein, Professor at the Department of Linguistics, UCLA.   Dr. Goldstein will speak on "Divergence-time estimation in Indo-European: The case of Latin".

Abstract:

Divergence-time estimation is one of the most important endeavors in historical linguistics. Its importance is matched only by its difficulty.  As Bayesian methods have become more common over the past two decades, two critical issues have come to the fore.

The first is that studies of divergence-time estimation have focused preponderately on root ages. Estimates of root ages rely crucially on the timing of clade-level events, but these events, despite being more recent, are often uncertain or even disputed. To put divergence-time estimates on more secure ground, more work at the clade level is essential.

The second problem is model sensitivity (e.g., Bromham 2019, Heggarty 2021). Researchers can now select from an array of tree, clock, and character-change models, the choice of which can have a dramatic impact on the estimated divergence times. The question then arises of how to choose among the bounty of modeling options.

My talk addresses these two issues in an investigation of a particularly fraught case within Indo-European, the diversification of Latin into the Romance languages. The results of this study support a gradualist account of the formation of the Romance languages that most likely began between 250 and 500 CE. It also offers strong support (with a posterior probability over 80%) for the view that Classical Latin is a sampled ancestor of the Romance languages.

At a more general level, I demonstrate that problems of model specification can be mitigated at least to an extent with increased model exploration and tests of model sensitivity and adequacy.

About Dr. Goldstein:

David holds a joint position in the Department of Linguistics and Program in Indo-European Studies at UCLA as well as a courtesy appointment in the Department of Classics.

His research is devoted to two broad areas.

• The first is the relationship between language change and linguistic theory, with a particular focus on syntactic and morphosyntactic change in Indo-European.
• The second is computational phylogenetics. 

From December 2022 through July 2023, David will be a Visiting Fellow at Clare Hall, University of Cambridge. During Easter Term 2023, he will be a Lewis-Gibson Fellow at the Cambridge Centre for Greek Studies.

28th October 2022 12:20 PM

Phonetics Lab Meeting

(optional for dissertators)

We will watch this video presentation by Shain & Elsner from a 2019 ACL Conference : https://aclanthology.org/N19-1007/, in preparation for our reading of their paper: 

Measuring the Perceptual Availability of Phonological Features during Language Acquisition using Unsupervised binary Stochastic Autoencoders

2nd November 2022 12:20 PM

PhonDAWG - Phonetics Lab Data Analysis Working Group

Sam will show the calculations behind the H, C, and V-measures for evaluating clustering, and we'll talk about what these measures mean.

3rd November 2022 04:30 PM

Linguistics Colloquium Speaker: Matt Wagers

The Department of Linguistics proudly presents Dr. Matthew Wagers, Professor and Department Chair at the Department of Linguistics, UC Santa Cruz.  

 Dr. Wagers will speak on "Alignment, Reanalysis and Reencoding in a Verb-initial Language".

Talk Abstract:

When do two DPs interact in sentence processing, and how does their similarity relationship affect that interaction?

My principal case study will be the processing of relative clauses (RCs) in Santiago Laxopa Zapotec (SLZ), an Oto-Manguean language of Oaxaca. SLZ is a verb-initial language whose ambiguous RCs consist of N - V - N sequences. Resumptive pronouns (RPs) can be inserted to disambiguate, either in subject, or object position; but object RPs sometimes appear to be quite difficult to comprehend.

In a series of judgment and eye-tracking experiments, we find that SLZ -- like many languages -- shows a version of the Subject Advantage. However, it is weak and defeasible.

We further find that the similarity of the two DPs is a principal determinant of difficulty. We take this as evidence of encoding interference, a somewhat surprising finding given the positional distinctness of the two nouns. We link the existence of encoding interference, instead, to their interaction via reanalysis.

Finally, the results from SLZ will be compared to similar, recent findings in other languages with optional object RPs (Hebrew) and ambiguous N - V - N RCs (Chamorro).

Bio:

Dr. Wagers is a linguist at the University of California, Santa Cruz. His research interest is in language processing, especially the coordination of syntactic information in memory.  He teaches courses at UCSC on psycholinguistics, experimental methods & design, language and memory, and syntax.

Dr. Wagers completed his Ph.D. in Linguistics at UMD (2008), supervised by Colin Phillips. Before that, he received his A.B. in Molecular Biology at Princeton University (2003), with Certificates in Neuroscience and in Linguistics.