Colloquium Fall 2018


University of Arizona

LING 495A/ 595A – Linguistics Department Colloquium

Fall 2018

Organizer:      Masha Fedzechkina & Andy Wedel

Schedule:      Fridays, 3:00pm – 4:30pm in Communication 311


September 14 NO COLLOQUIUM
September 21 NO COLLOQUIUM
September 28 Ken Forster (University of Arizona)
October 12: Samira Farwaneh (School of Middle Eastern & North African Studies, University of Arizona)
October 19: Special two-speaker Linguistics colloquium event on Friday October 19, 3-5 pm, Comm 311
Caleb Everett, University of Miami (3-4 pm):
Ease of articulation as a shaper of speech: The word list evidence
In this talk I discuss new findings on the extent to which ease of articulation impacts speech. The findings follow from an analysis of the rates at which numerous consonants occur in transcribed word lists representing thousands of languages. For instance, the data suggest that obstruents produced with oral obstruction closer to the glottis are less likely to be voiced, when contrasted with anterior obstruents. While this pattern is explainable via previously documented aerodynamic factors, the results presented suggest that such seemingly subtle factors have a more pervasive influence on speech than is typically assumed. This and other patterns are evident even after controlling for the relatedness and areal proximity of language varieties. The patterns are consistent with previous typological research on phoneme inventories, but also suggest that analyses of phoneme inventories may underestimate the influence that some ease-related phenomena have on speech.
Claire Bowern, Yale University (4-5 pm): 
Language, Culture, and Australian Exceptionalism"
Much recent work in linguistics has considered Australian languages as 'exceptional' in various ways. While the earlier strong position of Dixon's -- that Australian languages are a special case of language where comparative methods do not apply -- is now thoroughly refuted, there remain a number of ways in which Australian languages appear to form a distinct group. For example, despite 6000 years of change from a common ancestor (Bouckaert et al, 2018), most Pama-Nyungan languages have close to identical correspondence sets. More than 80% have consistent, initial stress. And sound correspondences across stable vocabulary in Pama-Nyungan indicate surprisingly few examples of sound change compared to the amount of lexical replacement seen (though the changes we do see and can reconstruct show regular principles according to the comparative method). In this talk, I review possible lines of inquiry for how and why Australian languages (either individually or in the aggregate) might show these features. I argue that cultural (ie, multilingualism) or physiological (e.g. Otitis media) explanations are unlikely on their own to contribute much to our understanding, and instead we should be looking towards explanations that consider the systematic stability of linguistic features.
October 26: Special talk by Limor Raviv
November 2: Shannon Bischoff (Department of English and Linguistics, Purdue University Fort Wayne)
November 9: Julie Legate (Department of Linguistics, University of Pennsylvania)
TITLE: Possible Voices
ABSTRACT: In this talk, I start from the hypothesis that the only innate, domain-specific property of language is binary branching Merge (see for example Berwick & Chomsky 2016).  I investigate the range of possible VoicePs in human language, where VoiceP is a cover term for the projection associated with initiators and transitivity.  I demonstrate that, in accordance with the guiding hypothesis, the range of possible voices is broader than standardly assumed, while noting a few restrictions to be explained.
November 16: Vera Gribanova (Department of Linguistics, Stanford University)
TITLE: Predicate formation strategies and verb-stranding ellipsis in Uzbek
ABSTRACT: In this talk, I present an investigation of the interaction between head movement and ellipsis of vP/AspP/TP (verb-stranding ellipsis, VSE) in Uzbek — an under-studied, head-final Turkic language of Central Asia. I argue that Uzbek has two distinct strategies for the composition of predicates: one is head movement (for verbal predicates) and the other is merger under adjacency (for non-verbal predicates). Since the independently demonstrable availability of head movement is a requirement for verb-stranding ellipsis, this makes the prediction that VSE could in theory be available for verbal, but never non-verbal predicates in Uzbek. I show that this prediction is borne out: non-verbal predicates permit only argument ellipsis while verbal predicates permit both argument ellipsis and VSE. In configurations involving Uzbek finite verbs, certain environments — those with object depictives and those with adjectival complements — permit only verb-stranding ellipsis but never argument ellipsis. In just this set of environments, Uzbek VSE imposes a strict identity requirement on the material extracted from the ellipsis site.
This last observation tells us a lot about the architectural status of head movement and its interaction with ellipsis. Syntactic movement of phrases out of ellipsis sites (e.g. in sluicing) permits violations of lexical identity of the extracted material under focus (Schuyler, 2001; Merchant, 2001). This is usually attributed to the syntactic nature of the movement and to the licensing condition on ellipsis (Rooth 1992, Heim 1997, Merchant 2001), which takes distinct variables inside the ellipsis domain and its antecedent to be identical. For head movement out of ellipsis sites, the results are variable across languages. Languages like Russian — among them Hungarian, European Portuguese, and Swahili — permit mismatches between extracted parts of the verbal complex and their corresponding antecedent components under focus, just as with phrasal extraction in sluicing. Languages like Irish and Scottish Gaelic do not permit such mismatches under any circumstances, potentially pointing to a postsyntactic status for head movement: if there is no syntactic movement out of the ellipsis site, this gives rise to a total identity requirement (Schoorlemmer and Temmerman, 2012; McCloskey, 2016). Recent studies (Landau to appear; Thoms 2018; Merchant 2018) have raised the question of whether the Irish and Scottish Gaelic pattern may be a language-specific anomaly, explainable by other means. If this is correct, then the crosslinguistic behavior of head movement out of ellipsis sites is not really variable, and is just like that of phrasal movement, as in the Russian-type languages.
The observation that genuine VSE in Uzbek imposes a strict identity requirement on material that is head-moved out of the ellipsis site supports the argument that an absolute identity requirement is genuinely present in a certain set of (typologically diverse) languages (though not all, as exemplified by Russian VSE). As Gribanova (2018) points out, this variable crosslinguistic behavior when it comes to identity requirements on the element(s) that are head-moved out of an ellipsis site is predicted
by an independently supported view of head movement (Harizanov and Gribanova 2018) in which certain types of head movement are syntactic (yielding the Russian-type VSE pattern), while others are postsyntactic (yielding the Irish/Uzbek-type VSE pattern). The Uzbek investigation therefore provides crucial evidence in favor of a particular view of the crosslinguistic landscape of VSE, and moves us a step closer to explaining why head movement out of ellipsis domains varies systematically in its behavior across languages.
November 23: NO COLLOQUIUM
November 30: Gabriela Caballero (Department of Linguistics, University of California San Diego)
TITLE: The interaction between lexical and grammatical tone in Choguita Rarámuri (Tarahumara)*
ABSTRACT: The cross-linguistic study of tone has largely focused on its lexical phonological properties, its phonetic implementation and interaction with other prosodic phenomena, but the morphological role of tone is still under-documented: What kind of morphological information may tone convey across languages? And what mechanisms regulate the outcome when there are lexical and grammatical tones in conflict? This talk addresses these questions through the lens of Choguita Rarámuri (CR; Uto-Aztecan), a prosodically complex language of northern Mexico.
CR has three lexical tones exclusively realized in stressed syllables. Stress-accent is morphologically conditioned and tonal patterns are partially predictable from stress. Yet there is evidence for grammatical tone and tonal classes independent of stress. We argue that the full range of grammatical tone patterns in CR follows from an analysis that incorporates tonal underspecification and construction-specific tonal patterns as output-oriented schemas. This analysis captures several properties of this system, including: (i) arbitrary relationship between tone patterns of related forms, (ii) heterogeneous nature of morphosyntactic classes expressed by tone melodies, and (iii) overwriting/avoidance of lexical tone by grammatical tone. We contrast this analysis with an alternative morphemic analysis that has been proposed in the literature (Spahr 2016), and argue that a construction-based analysis makes the correct empirical predictions.
* Work in collaboration with Austin German (UCSD/UT Austin)