重音与声调的语言节律共性

以＂轻重音＂和＂声调＂来对语言归类的主流概念，似乎忽略了两者相同并互动的语音和音系性质。本文把轻重音与声调的语音性质（即：音高、音强、音长）结合，再延伸至两者在历时音变上的密切关系，乃至在音系上的异同，归纳出＂节律元素假想＂。此假想把声调和轻重音皆视为节律（亦作＂韵律＂）的体现，使语言能透过[CONTOUR]和[PITCH]两个元素做出分类，组合出四种节律类型。这个崭新的概念捕捉到轻重音与声调的共性，不仅提供了一个统一两者的节律理论框架，而且容纳了各种难以归类的语言现象。

The conventional way of classifying languages using stress and tone appears to have overlooked their shared and interactive phonetic and phonological properties. The paper explores the overlapping phonetic correlates( i.e. length, intensity and pitch) of tone and stress, the historical phonological evidence of the close relationship between the two, followed by a delineation of their phonological similarities and differences. Phonetically, it can be shown that tone, like stress, relies on cues from intensity and length, so pitch is not the only factor. This can be shown experimentally by deliberately suppressing pitch information through whispering of tonal contrasts to reveal their contrasts for length and intensity in comparison with normally phonated articulations. Similarly, there are pitch accent systems that use pitch to indicate accented syllables. Phonologically, while lexical tone is often stipulated in the lexicon, it must too be remembered that stress languages do allow specification of the locus of primary stress before the secondary stresses are derived by metrical principles. Further, stresses and tones both figure into the same consideration when used as qualification for prosodic wordhood: a minimal prosodic word must be bimoraic, which translates into at least one stress bearing mora/syllable for stress languages or one fully-toned mora/syllable for tone languages. Even in terms of spreading, tone and stress show overlap. While the conventional wisdom is that tones may spread as autosegments, one should be aware that the pitch correlate of stress too can spread, albeit only progressively. Historical studies reveal also an intimate relationship between tone and stress. There is evidence that the pitch accent system in Japanese evolved out of tone in proto-Japanese in ways similar to tonal systems found in Bantu languages.Further,there are modern tonal languages that can be demonstrated to have roots in stress accent systems. With the blurring of the distinction between tone and stress, it is nonetheless necessary to recognize the traditional motivations for distinguishing them. Specifically, traditional wisdom appeals to pitch during normal phonation and the presence of contour in the pitch melodies. Teasing out these two parameters, it becomes possible to reimagine prosody as parametric taxonomies, which is the bedrock of the "Prosodic Essence Conjecture" proposed in this paper. By treating both stress and tone to be the manifestation of prosody, the two can be collapsed into the same analytical umbrella. By appealing to distinctions in the specification of [contour] and [pitch], the framework yields a language typology of four: (i) [pitch, contour], where a language uses contoured pitch melodies in its prosody, e.g. Standard Chinese; (ii) [pitch], where a language uses only level pitch melodies, e.g. stress accent languages like Japanese; (iii) [contour], where a language does not employ pitch phonologically but uses other phonological contours such as diphthongized nucleus, e.g. English; and (iv) [?] where prosody is flat, e.g. Hawaiian. Using only two discrete parameters, the conjecture derives what appears to be gradation in the shades of prosodic taxonomy.