Cross-Linguistic Phonological Variation in Tone Systems: A Cross-Modal Analysis of Tonal Perception and Production in Multilingual Speakers

Project Abstract

This study investigates how tonal systems vary across languages and how multilingual speakers perceive and produce tone across modalities, integrating phonology, psycholinguistics, and cognitive neuroscience to illuminate cross-linguistic patterns of tonal representation and processing. We address how tonal contrasts—including contour tones, pitch height, and tonal timing—are learned, maintained, and generalized in multilingual contexts where native and non-native tonal inventories interact. The research employs a cross-modal framework, combining auditory production tasks, perceptual discrimination experiments, articulatory measurements (e.g., ultrasound tongue imaging, laryngeal videography, and high-speed kinematic data), and neural correlates captured via EEG and fMRI during tone perception and production. We hypothesize that multilinguals develop shared, abstract tonal representations that can be flexibly accessed across languages, yet show language-specific tuning driven by phonotactic constraints, prosodic context, and sociolinguistic factors such as age of acquisition and literacy. The study unfolds in three interconnected strands (i) cross-linguistic perception studies that quantify perceptual salience and categorical boundaries of tones in languages with contrasting tone inventories; (ii) production studies that examine tonal realization, coarticulation, and gestural alignment in spontaneous speech and elicited tasks, mapping how listeners infer intended meaning from tonal cues; and (iii) neurocognitive investigations that track processing efficiency, hemispheric specialization, and neural plasticity associated with managing multiple tonal systems. The methodology leverages a large, diverse corpus of multilingual speakers spanning language families with varied tonal typologies, complemented by a controlled laboratory protocol with matched stimuli across languages to isolate perceptual and production biases. Advanced statistical models, including mixed-effects regression, Bayesian hierarchical modeling, and machine learning classifiers, will parse the contributions of language history, proficiency, and social factors on tonal competence. Anticipated findings include evidence of cross-language transfer effects in both perception and production, with partial generalization of tonal categorization across languages sharing phonological features, and distinct modality-specific processing advantages for tones embedded in lexical versus grammatical functions. We expect that multimodal data will reveal convergent evidence for a shared prosodic architecture that supports flexible mapping between acoustic signals and phonological categories, while also indicating language-specific refinements shaped by exposure and usage. The implications extend to theories of phonological representation, second-language pedagogy for tone languages, and practical applications in speech technology, where robust cross-linguistic tonal models can enhance recognition, synthesis, and language-learning interfaces for multilingual populations. This integrated approach advances understanding of how humans manage complex tonal systems in real-world multilingual discourse and contributes to a comprehensive model of tone as a dynamic, modulated feature of spoken language.

Project Overview

What This Project Is About

---

This project looks at how tone systems in language (pitch patterns that change meaning) vary across languages and how people perceive and produce these tones in real life. It compares how speakers notice tonal differences with how they say them, across different language backgrounds, to see what links perception and production.

---

---

The Problem It Addresses

---

Many studies treat tone as fixed within a language, but in the real world people learn multiple languages with different tone rules. The gap is understanding how multilingual speakers handle these rules in both hearing and speaking, and what factors (like language family or exposure) change this process.

---

---

Objectives of the Project

---

1. Identify common patterns in how tones are perceived across several languages.
2. Examine how speakers produce tones when switching between languages with different tone systems.
3. Explore whether perception accuracy predicts production accuracy in multilinguals.
4. Suggest practical implications for language learning and speech technology.

---

---

What You Will Do Step by Step

---

1) Review basic literature on tone perception and production. 2) Recruit multilingual participants and design simple perception and production tasks. 3) Collect listening and speaking data in controlled and real-life contexts. 4) Analyze results to find links between perception and production. 5) Compare across language backgrounds and summarize findings.

---

---

---

---

Expected Outcome

Expect to show how tonal perception and production align or diverge for multilingual speakers, identify factors that influence this relationship, and offer ideas for teaching tones and improving speech software for multilingual users.