In the field of Second Language Acquisition (SLA), pedagogical and research focus has traditionally centered on "segmentals"—the individual vowels and consonants that form the building blocks of speech. However, this position paper, originally presented at INTERSPEECH 2009, argues for a critical paradigm shift toward the study of "suprasegmentals." These features, which include rhythm, stress, and intonation, constitute the "music" of a language and are often the primary determinants of a speaker's intelligibility and perceived proficiency.
The Challenge of Suprasegmentals
While segmentals are relatively discrete and easier to categorize, suprasegmentals are notoriously elusive. They are highly variable, influenced by a speaker’s emotional state, the specific lexical choices, and the broader discourse context. Furthermore, the field lacks a universal consensus on how to objectively measure the pragmatic meaning of a pitch change or a rhythmic pause.
This paper addresses the need for a structured, large-scale data collection effort to move beyond anecdotal observations. By establishing a systematic model of "interlanguage" prosody, researchers can better understand how a learner's native language (L1) habits are transferred into their second language (L2) production.
The AESOP Framework
The authors propose a methodology for the Asian English Speech cOrpus Project (AESOP), a multi-national collaborative effort involving researchers from Hong Kong, Taiwan, Japan, and Thailand. Unlike traditional corpora that might record spontaneous or random speech, AESOP utilizes designed "textual environments" to elicit specific communicative functions:
- Lexical Stress Elicitation: Using word pairs where stress placement determines the part of speech (e.g., 'record as a noun vs. re'cord as a verb) to see if learners can navigate English's mobile stress patterns.
- Utterance-level Focus: Prompts designed to force "narrow focus" (e.g., emphasizing a specific word to answer a targeted question), testing whether learners can use prosody to highlight new information.
- Syntactic Disambiguation: Using sentences where intonation alone changes the meaning, such as distinguishing between addressing a person ("The fight is over, Fred") and describing a situation ("The fight is over Fred").
- Speech Act Marking: Analyzing how learners use pitch contours to distinguish between statements and various types of questions (Wh-questions vs. Yes-No questions).
Cross-Linguistic Insights and Language Transfer
The paper synthesizes existing research to highlight how different L1 backgrounds create unique challenges for Asian learners of English:
- The Rhythm Gap: Many Asian languages, such as Chinese, are "syllable-timed," where each syllable has roughly equal duration. English, conversely, is "stress-timed." Chinese learners often apply their native rhythmic rules to English, resulting in a staccato rhythm that can be difficult for native listeners to parse.
- Pitch as a Primary Cue: Mandarin speakers often rely on pitch as the primary cue for English stress—likely a transfer from the tonal nature of their L1. In contrast, native English speakers use a complex combination of pitch, duration, intensity, and vowel quality.
- Mora-timed Interference: Japanese learners, coming from a "mora-timed" language, frequently struggle with vowel reduction in unstressed syllables. Because vowel length is a distinct linguistic marker in Japanese, they may resist shortening vowels in the way English requires for natural-sounding prosody.
Conclusion and Future Directions
As a position paper, the primary contribution is the establishment of a rigorous framework for the AESOP project. The authors acknowledge significant hurdles, particularly the difficulty of mapping specific acoustic features (like a 50Hz pitch rise) to a single underlying meaning.
The ultimate goal of this corpus-based approach is the development of advanced Computer-Aided Language Learning (CALL) systems. By understanding the systematic errors made by specific L1 groups, technology can be designed to provide corrective feedback that goes beyond identifying an error, explaining why the prosody failed (e.g., "You failed to reduce the vowel in the unstressed syllable"). This roadmap seeks to bridge the gap between theoretical linguistics and practical pedagogical tools for the millions of English learners across Asia.
Summary updated