Andrew Nevins
Encoding and decoding in the whistled phonology of Antia, Greece
This is intended to be a discussion about the phonetics and phonology of whistled languages, and participants are invited to read the attached paper by Annie Rialland on the topic as a starting point. I will also present some production data from words and non-words collected in Antia, the results of perception tasks with both whistlers and non-whistling Greek speakers, and offer some ideas about the encoding mechanism used in this surrogate speech system.
Recommended reading: Rialland (2005) Phonological and phonetic aspects of whistled languages
Peter Graff
Evolutionary vs. Phonetically Driven Phonology: An Iterative Learning Experiment
In this talk, I will propose an iterative learning experiment trying to test the predictions of Evolutionary Phonology (Blevins, 2006) and Phonetically Driven Phonology (Hayes and Steriade, 2004). Both hypotheses about phonological learnability and knowledge predict phonological systems to be optimized for transmission. The crucial difference is that Phonetically Driven Phonology hypothesizes phonetic optimization of phonology to be speaker driven, while Evolutionary Phonology attributes phonetic optimality to unbiased or “innocent” misperception and production independent of the grammar of the speaker. I will suggest simulating diachronic transmission in iterative learning and propose ways in which to manipulate speaker driven optimization to see whether such manipulation affects the course of simulated linguistic history as might be predicted by certain conceptions of Phonetically Driven Phonology.
Youngah Do
Child Preference of Base Correspondence: the Asymmetry of the Inflection of Regular and Irregular Verbs in Korean
In this talk, I examine the distinctive ways in which children inflect Korean verbs with respect to the (ir)regularity of the verbal stem. An experiment of picture description asks children to inflect two verb forms in a coordinated sentence. The result shows that the inflectional structure of the two coordinated verbs are always identical. Interestingly, an asymmetry of the inflection is found according to the order of the appearance of regular and irregular verbs in a sentence. When the first verbal position is occupied by regular verb and irregular verb follows, children inflect both verbs in a simple way(C category stem+C category suffix), not using any extra morpheme. On the contrary, when the inflection of irregular verb is required first and regular one is following, they inflect the verbs in a complex way by using an extra morpheme (A category stem+A category suffix+ Extra morpheme +C category suffix).Adopting the hypothesis that that the A category is the base in the Korean verbal inflectional paradigm (Albright and Kang 2009), I argue that this asymmetry is due to children’s tendency for respecting base correspondence in the process of verbal inflection.
Sverre Johnsen (Harvard)
Contrast maintenance effects in Norwegian retroflexion
Norwegian contrasts morpheme initial /s/ and /?/ before a vowel (/sV-/ - /?V/), but not before a consonant (/sC-/ - */?C-/). Initial /s/ undergoes retroflexion to /?/ when preceded by a morpheme ending in /-r/. Before a vowel, the contrast between /s/ and /?/ is then lost (/sV-/, /?V-/ > /?V-/), but no contrast is lost before a consonant (/sC-/ > /?C-/). I present data from two experiments showing that Norwegian speakers apply this retroflexion significantly more often before a consonant than before a vowel. The effect is that a contrast is better maintained. Retroflexed /?V/-tokens have, however, more lexical neighbors and a higher phonotactic probability than retroflexed /?C/-tokens. Experiments in lexical access have shown that items with a high lexical neighborhood and phonotactic probability are more prone to be misidentified. I show that under an exemplar model of speech perception/production, the asymmetry in retroflexion between /sV-/ and /sC-/ is a direct consequence of such misidentifications.
Michael Tanenhaus
Fine-grained phonetic detail in spoken word recognition
Despite considerable evidence to the contrary, it is widely assumed that some classes of speech sounds are perceived categorically in a way that exemplars from other types of non-speech categories are not. Yet, the articulation of many sounds, including consonants, varies systematically with position in a prosodic domain. A system that discarded sub-phonetic detail would thus be ignoring potentially useful information. I’ll review recent data from eye-tracking studies demonstrating that spoken word recognition does, in fact, exploit fine-grained sub-phonetic detail to make probabilistic hypothesis about lexical candidates, including within-category variation for stop consonants—the poster child for categorical perception.
