Tone is one of the common features of the world's languages. It is found not only in East and Southeast Asian languages, but in many languages of the world: native American, New Guinean, African, etc. If we add to this category pitch-accent languages, an overwhelming number of the world's languages can be seen to exploit this particular phonetic feature in varying degrees. Four languages and language groups are chosen in this issue to exemplify various tone types; namely, Chinese, Korean, Haida (North American) and Bantu (Africa). Although typology is considered in the analysis of these languages, we mainly intend to show different tone and/or pitch-accent languages, no unification being attempted as to the usage of the terms tone and pitch-accent.
This paper describes some of the typical pitch accent systems found in Korean dialects and discusses various tonal or accentual phenomena that are important for a deeper understanding of each of the pitch systems. Typological comparisons with the pitch accent systems of Japanese are also made. The author also gives a provisional classification of all the Korean pitch accent systems so far reported, including non-distinctive ones, based on several synchronic criteria.
The monosyllabic lexical tone in Chinese is a composite of the two prosodic components that interact with each other as constraints. “Tonal features” specifically refer to the features associated with pitch, including register and contour, while the “non-tonal features” refer to those related to phonation types, including consonant types, vowel quality and syllable duration. The stress effect on the tonal features in polysyllabic words is studied. In the dialects attested, two stress patterns are coexistent: right stressed pattern and left stressed pattern. Stress reduction could result in the neutralization of tonal features, which varies in manner and degree across the dialects. Four types of neutralization are observed in the dialects: (α) Context dependent, local type, (β) Context dependent, systematic type, (γ) Context free, partial type and (δ) Context free, complete type. Type (δ) is the case of the so called “neutral tone”, which forms a “word tone” in combination with the tone in the stressed syllable. An ultimate consequence of tonal neutralization is monotonalization, which nowadays is going on in the new types of the Wu dialects, such as those in Shanghai and in Hangzhou. An interesting finding is that a sort of accentual system emerges in the process of monotonalization. Some related topics are discussed together with the historical background of the development of word tones in northern dialects.
Haida, a Native American language spoken in the northwestern coast area of British Columbia in Canada, has three levels of phonetic tone, namely high (H), mid (M), and low (L), whose occurrences correlate with the syllable structure and the morphological process. The present study observes distributions of each tone in monosyllabic and polysyllabic words and argues that occurrences of tones are not specified in the underlying level but are predictable from the syllable structure and certain morphological information. Based on this fact, it postulates a set of tone assignment rules for deriving these phonetic tones.
The Bantu languages of Africa show various tone types, ranging from pure tone languages to non-tonal languages. In this paper the author analyses the tone system of nouns in two languges: Tembo (J.57, eastern Congo) and Haya (J.22, northwestern Tanzania). With Tembo it is demonstrated that in this language each mora in principle should be marked either as high, low, falling or rising (the last two being combinations of high and low). The number of partterns increases in geometric progression. Tembo, therefore, can be characterized as a polymoraic tone language. In Haya, however, the number of tone patterns increases in arithmetic progression, as a function of the number of the syllables in the stem. High tone has become accentual in nature in Haya, in which only one high tone syllable per word is permitted in isolation. For typological reasons, examples of Safwa (M.25) and Swahili (G.42) are added to illustrate a language in which the number of patterns remains stable regardless of the number of the syllables (Safwa), and a bound accent language (Swahili).
This paper is concerned with the effect of accentual fall on final lengthening in Japanese. A series of experiments have revealed that in isolated utterances the final vowels of words with no accentual fall are significantly longer by about 40ms than those with accentual fall. Also, in phrases composed of a noun and a particle /no/ plus phrase-final noun, a similar amout of final lengthening is observed the phrase-final nouns with no accentual fall, but not for their counterparts with accentual fall. These results suggest that unaccented or finally-accented words have their final vowels lengthened in the utterance final position as a means of marking the end of utterances.
The present study investigated how monolingual speakers of Japanese and English and bilingual Japanese speakers of English were sensitive to syllables and morae in the two languages. They were presented with spoken words in both languages and asked to mark on a written transcript of each word (e.g., buranko for Japanese and veranda for English) the second natural division point from the onset of the word. The statistical analysis showed that both the Japanese and English monolinguals were sensitive to morae and syllables, respectively, and that the bilinguals were sensitive to syllables in English but neither to syllables nor to morae in Japanese. These results suggest that while monolinguals may be sensitive to a single unit, bilinguals may have access to plural units in the mental lexicon.