Part 1: 10 am–12 pm, Friday, October 2, 2020 [JST: Japan Standard Time]
Kathryn Franich (U Delaware)
"Rhythmic Alignment in Medʉmba Language and Song: The Role of Metrical Foot Structure"
Across many different languages and musical traditions of the world, metrical structure has been found to play a role in the rhythmic mapping of speech to music. Many African tone languages lack clear evidence of metrical stress, despite often showing phonological patterns which are consistent with the existence of foot structure. Whether feet are truly rhythmic in nature in these languages remains a question of some debate, however. This talk presents evidence from Medʉmba, a Grassfields Bantu language, that foot heads are, indeed, rhythmically prominent in the language. I draw on evidence from a speech production task as well as from text-setting patterns in two traditional songs that foot heads, which tend to align with the left edge of word stems, play a role in rhythmic coordination akin to stressed syllables in languages like English. I show that, while the precise alignment of metrically-prominent syllables in music is culturally-dependent, distributional patterns nonetheless evidence a common rhythmic function of metrically-prominent syllables in music across languages and cultures.
Laura McPherson (U Dartmouth)
"Grammatical tone and its segmental correlates: Insights for analysis"
About half of the world's languages are tonal, i.e. pitch is used to differentiate lexical meaning of at least some morphemes. While this lexical tone receives the most attention and is the most familiar, the vast majority of tone languages also display "grammatical tone", i.e. pitch used or triggered by the morphosyntax. Despite its prevalence, grammatical tone remains understudied and undertheorized. This talk provides a general overview of the varied forms that grammatical tone can take, from clear tonal morphemes to tone changes triggered by syntactic structure without seeming to contribute any independent meaning. Parallels are drawn between these kinds of grammatical tone and better known phenomena in segmental morphology, serving both to demystify grammatical tone and to suggest avenues for theoretical analysis. I show that the category "grammatical tone" encompasses a diverse set of phenomena and defies any unitary analysis. Like Hyman (2007) famously put it, tone can do everything segmental morphology and phonology can do---and more.
Part 2: 5–7 pm, Friday, October 16, 2020 [JST]
Grammar shaped by melody – a tune-driven pathway to morphology
Timo Roettger (U Osnabrück)
"Human speech has multiple information channels that can simultaneously signal different levels of meaning. One such channel consists of consonants and vowels organized into metrical structures, referred to as the text. Largely independent of these structures, there are suprasegmental modulations superimposed on the text, most notably intonation, referred to as the tune.
The text and the tune are often conceived of as separate levels of phonological representation. These levels, however, exhibit a fundamental inter-dependence which is rooted in the nature of the speech transmission process. For intonation to be produced, voiced segmental material is needed to enable the vocal folds to vibrate; for intonation to be perceived, it is important for the segmental material to have periodic energy with a rich harmonic structure
I will show that when segmental material does not allow for vocal fold modulation, or is not sufficiently sonorous to allow for optimal perceptual retrieval of pitch, languages adjust the text to provide an adequate carrier signal for intonational meaning. I will present evidence from a wide variety of languages which exhibit prosodically conditioned insertion of non-lexical vowels, suppression of otherwise regular vowel devoicing, or lengthening of existing vowels in order to accommodate the requirement to realize communicatively relevant tones. I will further show that this variation might even be morphologized in some contexts. I will discuss vocative constructions which commonly co-occur with particular tunes. Using a cross-linguistic corpus, I show that vocative morphology is particularly tune-friendly.
The available evidence suggest that intonation poses functional pressure on its segmental hosts, leading to the temporal adjustment and preservation of existing elements or the insertion of new elements, so as to ensure the realization of intonational events. I will argue that these interactions can lead to re-occurring segmental alternations and shape grammatical morphemes."
Derived-environment effects: a view from learning
Adam Chong (Queen Mary University of London)
Morphologically derived-environment effects (MDEEs) are well-known examples in which static phonotactic patterns in the lexicon do not accord with what is allowed at morphological boundaries (phonological alternations). In this talk, I examine these patterns from two complementary perspectives. First, I consider what consequences these patterns have for phonological learning. Using an artificial grammar learning experiment, I show that alternation learning involving MDEEs is impeded. This provides support for current models of phonological learning in which the phonotactic learning supports alternation learning. I then present corpus and phonotactic modeling results examining the status of the phonotactic patterns in the lexicon of two well-known MDEE cases: Korean palatalization and Turkish Velar Deletion. I show that in both languages the reported mismatches between phonotactics and alternations are superficial. This undermines existing analytic assumptions in previous analyses of these patterns. I discuss the implications of these findings for a theory of MDEEs, as well as for the relationship between phonotactics and alternations in learning and the grammar.
Plenary Talk: 9–11 am, Saturday, October 31, 2020 [JST]
Donca Steriade (MIT) "Cyclicity Generalized"
Co-hosted by the Phonetic Society of Japan
Abstract [LINK]
Part 3: 10 am–12 pm, Friday, November 13, 2020 [JST]
Oriana Kilbourn-Ceron (Northwestern U)
Phonological variation in connected speech as a window on speech production planning (and vice versa)
Words in connected speech can differ substantially from their “citation” forms, and it is often difficult to predict exactly how they will be pronounced with complete certainty. One thing that is clear is that phonological context plays an important role, but does not always fully determine the outcome, leading to phonological variation — a situation where multiple phonetic outcomes are possible within the same context. My research investigates the use of context-specific allophones (e.g., flap in English, devoiced high vowels in Japanese) in relation to speech production planning. I propose that limitations on advance planning during word form encoding have consequences for patterns of phonological variation. I will present results from both corpus and experimental studies which show that context-specific allophones are less likely to be used when the upcoming word is more difficult to plan, which limits the degree of advance planning and blocks the availability of upcoming phonological contexts. This provides both a new insight into patterns of phonological variation, and a new tool for investigating the extent of advance planning during speech production.
Samuel Tilsen (Cornell U)
The emergence of non-local phonological patterns in the Selection-coordination-intention framework
In this talk I will present some of my recent ideas on how non-local phonological patterns can emerge in a gestural framework (see Tilsen 2019, Motoric mechanisms for the emergence of non-local phonological patterns). Some patterns—e.g. vowel harmonies, nasal harmonies—can be readily analyzed to arise from temporal extension of articulatory gestures (i.e. spreading); such patterns can be viewed as articulatorily local. However, there are other patterns—e.g. nasal consonant harmony, laryngeal feature harmony—which cannot be analyzed as spreading; instead these patterns appear to enforce agreement between features of similar segments without affecting intervening segments, an apparent form of action-at-a-distance. Indeed, there are numerous typological differences between spreading harmonies and agreement harmonies. This suggests that there is a mechanistic difference in the ways that spreading and agreement harmonies arise. I argue that in order to satisfactorily understand spreading and agreement patterns, the gestural framework of Articulatory Phonology must be enriched with respect to how targets of the vocal tract are controlled in planning and production. Specifically, it is proposed that production models should distinguish between excitatory and inhibitory articulatory gestures, and that gestures which are below a selection threshold can influence the state of the vocal tract, despite not being active. This idea is applied to provide two distinct mechanisms for the emergence of non-local phonological patterns.
Part 4: 10 am – 12 pm, Friday, November 27, 2020 [JST]
Chikako Takahashi (Stony Brook U)
Phonetic interaction between late Japanese-English bilinguals’ L1 and L2 vowels
In this talk, I will present my work on the phonetic interaction between late Japanese-English bilinguals’ L1 and L2 vowels, where I investigate how learning a new L2 English vowel influences the perception and production of their L1 vowels. Specifically I focus on how production and perception of their L1 vowels are influenced by relative L1:L2 dominance factors as well as their L2 production and perceptual ability. The study recruited 65 late L1-Japanese learners of L2 English whose L1 Japanese and L2 English perception and production of target vowels were compared to monolingual control groups (Japanese and English). Our data show L1 vowel drift in both perception and production, supporting the view that late L2 learners’ L1:L2 phonetic systems are dynamic and malleable. We identified L2 production acuity as a major factor associated with L1 assimilatory drift and also demonstrated a complex interplay between L2 production acuity, L2 perceptual ability, and L2 dominance that together appear to influence the nature of the L1 drift.
Jianjing Kuang (U Pennsylvania)
Tone acquisition beyond f0
Traditionally, tone is defined by a single dimension of the voice source: fundamental frequency, or F0. However, as demonstrated in my research, effective tone production and perception in fact integrates other articulatory and acoustic dimensions in the voice source (i.e. voice quality cues). Previous studies on tone acquisition and development mostly focus on f0 cues, and little is known how voice quality cues become integrated in the tone production and perception. In this talk, as the first step to answer this question, I will present evidence from computational modeling of tonal productions from Mandarin-speaking children and adults, showing that cue integration of f0 and voice quality cues in the tonal contrasts exhibits a developmental curve for Mandarin-speaking children. These findings support the multidimensional tone model proposed by Kuang (2013), and provide important new insights for tone acquisition.