Conflict and cooperation, or: how I learned to stop arguing and love interaction

Sunday, Nov. 3 | 9:00 - 10:00

Generative phonological analysis relies on interactions (ordering, ranking) between grammatical statements (rules, constraints). These interactions primarily serve to simplify rules or constraints by reducing redundancy across them, a result argued to lead to greater insight into the nature of phonology. Some simplifications are only achievable with rule ordering, while others require constraint ranking. Drawing on collaborative work with Lev Blumenfeld, this talk first maps out the formal typological landscape of process interactions with ordered rules. We then venture into new territory, beginning to explore the landscape of process interactions with ranked constraints. At the heart of both of these landscape explorations lies a fundamental question: how do different theoretical approaches illuminate distinct aspects of phonological patterns? By comparing the different simplification strategies available with rule ordering and constraint ranking, we gain insight into the unique analytical power of each framework. This talk thus advocates for a nuanced view of phonological theory — one that recognizes formal models not as absolute truths, but as sophisticated tools for uncovering the underlying principles of phonology.

Co-Speech Gestures as a Window into Phonological Knowledge

Saturday, Nov. 2 | 16:45 - 17:45

Phonological description and analysis tend to draw heavily on a combination of speaker intuitions and acoustic-phonetic output from speech as evidence for grammatical patterns. Some phenomena, such as stress and metrical structure, tend to be difficult to assess based on these types of data alone: speaker intuitions about metrical prominence are often difficult to elicit, and phonetic correlates to stress are highly variable across languages and easily confused with cues to other prosodic events (de Lacy 2012; Roettger & Gordon 2017). In this talk, I discuss how co-speech gestures—movements of the hands, arms, head, etc. that are temporally coordinated to speech—can highlight patterns of prominence which may be difficult to assess based on acoustic patterns alone. I draw on data from two Niger-Congo languages (Medʉmba and Igbo) with quite different prosodic structures to demonstrate how gesture provides a unifying source of evidence of prominence across the two. For Medʉmba, gestures tend to cluster around the initial position of the stem, while for Igbo, competing constraints on the location of gestures suggest metrical organization at both the word level and the tonal level. I then briefly sketch some links between gesture and music as sister ‘coordinative practices’ in revealing patterns of metrical organization in language. Finally, I present preliminary cross-linguistic comparisons of gesture timing in Medʉmba and English (US and Cameroonian varieties) which suggest that gesture (and coordinative practices more broadly) may have a role to play in explaining typological variation in the phonetic manifestation of metrical prominence.

Optimality Theory with lexical insertion is not computable
Friday, Nov. 1 | 16:45 - 17:45

This talk examines the computational consequences of introducing lexical insertion, i.e., the ability to copy morphemes or insert them from the lexicon, as an operation into Optimality Theory. I demonstrate that this operation makes OT not computable: in other words, it is impossible to determine the output of a given input in a finite amount of time. This result is derived by modeling the Post Correspondence Problem in an OT grammar that uses only representations and mechanisms attested in the literature.