Prosodic Contrasts and Segmental Analysis in Himalayan Languages

Abstract

The connection between pitch, phonation and plosive voicing has been observed in many of the languages of the world, including those of the Himalayas. The paper examines different analyses of largely phonetically similar data from various Tibetan and Tamangic languages, comparing and evaluating the substantially different analyses offered in different descriptive traditions, arguing that principles of economy and descriptive adequacy allow us to choose between alternative analyses for many languages that have been described with different types of contrasts involving plosive manner and pitch height.

Keywords:

• Tone
• Phonology
• Phonation
• Himalayas
• Tibetan
• Tibeto-Burman
• Bhutan
• Dzongkha

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1 Further complications in the form of contrasting tonal contour will not be addressed here, as they are not relevant to the discussion about the voicing/pitch correlation that is the focus here. Additional contours, either high or low, are frequently encountered in languages of the Himalayas.

2 Additionally, breathiness has been observed to correlate with low rising pitches.

3 It is precisely the contrast between different stops in the low register that appears to be diachronically unstable. The Tamang system described in Section 4 is one example of a language in which there is no VOT contrast in low register.

4 The description of ‘Central Tibetan’ holds for the dialects of Lhasa, Dingri, Shigatse, Drokpa, Yohlmo and others (e.g. Hari 1979, 2010; Tournadre & Dorje 2003; Kretschmar 1986; Herrmann 1989; Kretschmar 1986, 1995; Haller 2000; and others).

5 Van Driem (1995: 10) notes for Dzongkha that ‘In contrast to the normal voiceless stops, the secondary voiceless stops are followed by a breathy vowel in low tone’.

6 The transcription of reduced aspiration follows, e.g. Kretschmar: ‘Die Aspiration, die nur im absoluten Anlaut auftritt, ist bei hochtonigen Silben stark ausgeprägt, bei tieftonigen wird sie auch fakultativ schwächer realisiert’ (‘Aspiration, which only contrasts initially, is pronounced strongly in high-toned syllables, and is realised relatively weakly in low-toned syllables’) (Kretschmar 1986: 23). I use the same conventions as Kretschmar, with the contrastive transcriptions [ph] in high tone syllables, and [ph] ∼ [pʰ] in low tone syllables, for the underlying /pʰ/ series. Similarly, the unaspirated series varies in terms of VOT: ‘Verschlußlaute werden in tieftoniger Silbe fakultative mehr oder weniger gespannt realisiert’ (‘stops are realised as more or less voiced in low-toned syllables’), transcribed as [p̂] ∼ [p] (Kretschmar 1986: 23). Watters (2002) describes similar data in related Tibetan languages, though a phonemic, as opposed to phonetic, analysis is underdeveloped.

7 Note that these are not the same as principles, for instance, principles of dispersion or ease of articulation and enhancement (as articulated in, for instance, Liljencrants and Lindblom (1972), Quantal theory—Stevens (1989), Stevens and Keyser (1989), and others). They are phonological principles that operate in addition to these phonetic principles.

8 For instance, we might spuriously posit that dear and deer were distinguished underlyingly by the initial stops, /diːə/ ‘dear’ vs. /ɖiːə/ ‘deer’, a contrast which is neutralized to [d] in all environments.

9 An alternative, and equally unmotivated, treatment of the putative phonological contrast found in dear vs. deer might posit the underlying forms /diːə/ and (for instance) /ɢʟiːə/, or /ɢʟy/ (plus some rules or principles to convert these underlying forms to [diːə]).

10 Note that the contrast in phonation has not been counted as an underlying part of any of these analyses, being able to be derived from combinations of plosive manner contrasts and pitch height contrasts. This means, of course, that it would also be possible to analyse away the pitch height contrast in favour of phonation in at least some languages. The fact that most of the languages in question have pitch contour contrasts as well means that we must, at any rate, consider pitch to be underlying, and so for the sake of parsimony of variables the possible analyses that minimize underlying tonal differences have not been exemplified here.

11 The simple, and rather deterministic, explication presented here does not do justice to the more nuanced arguments that have been cogently presented by many Tibeto-Burmanists. The interested reader is referred to works by the authors cited at the end of this paper.

12 The derivation of the South Mustang data would follow essentially the same course as the Dzongkha given in examples (5) and (6), without mention of breathiness (since that is not described in the source).

13 Some varieties of Dzongkha are also reported to have a system that corresponds to the Tamang or Tshangla systems described here (thanks to an anonymous referee for this information).

14 While stating that ‘the tone of … the voiced syllables is in fact redundant, as these syllables are sufficiently distinguished by their onset consonants’, Huber (2005: 20) notes that ‘there are a few exceptional words with a voiced stop onset but a high register tone’. This emphasizes the need for detailed language description, allowing for a richer typological view of the variation found. Qualifications such as these have not been reported for the other languages surveyed here, where plosive VOT and pitch height are more consistently correlated.

15 A possible analysis involves splitting the pitch space into an upper and –upper register (following Yip 1989, 2002). We would describe the [H] and [M] tones as being [+upper, H] and [+upper, L], respectively, and the [L] tone as [–upper, L]. Breathiness is associated with the [–upper] register, where [voiceless aspirated] is not felicitous; equally, the upper register is not compatible in the main with [voiced] stops.

16 I am using ‘plosive’ in a not entirely traditional sense to include ejective and implosive stops, but one that captures diachronic relationships more than excluding these consonants would. The different manners attested in plosives are: voiceless, voiced, voiceless aspirated, ejective, implosive/glottalized, prenasalized (voiced or voiceless), postnasalized and breathy (secondary articulations such as rounding or palatalization are not included as different manners. Of these, aspiration and implosion correlate most strongly with tonality. ‘Eurasia’ is defined as in Hammarström and Donohue (2014).

17 For Africa (and no other continent), tonality correlates positively with the group of languages without manner contrasts in plosives, demonstrating that these trends are areally governed, and not an artefact of the overall phonological ‘complexity’.

18 The languages close to Bhutan in Figure 1 are, from the north clockwise: Tibetan (Lhasa, Shigatse, to the north), Tiwa, Koch (Wanang) (to the south-east and south), and two languages of Nepal to the west, Yakkha (Kiranti) and Lhomi (also Tibetan).