Historical transfer of nasality between consonantal onset and vowel: from c to V or from V to C? icon

Historical transfer of nasality between consonantal onset and vowel: from c to V or from V to C?

страницы:   1   2   3   4   5   6   7   8   9

Historical transfer of nasality between consonantal onset and vowel: from C to V or from V to C?

Alexis MICHAUD, Guillaume JACQUES and Robert L. RANKIN

Author manuscript. Forthcoming in Diachronica 29:2 (2012).

Summary: Comparative data from several language families show that nasality can be transferred between a syllable-initial consonant cluster and the following vowel. The cases reported to date are summarized, and a new analysis is proposed for a set of Sino-Tibetan data. The evolution appears to go in both directions: from the consonantal onset to the following vowel in Tai-Kadai, Austroasiatic, Sino-Tibetan, Niger-Congo (Kwa) and Indo-European (Celtic), and from the vowel to the preceding consonant in Siouan. However, an examination of the conditions on these changes brings out an asymmetry. In most cases, transfers of nasality take place from a consonantal onset to a following vowel; the instances we found of a regular change in the opposite direction all come from languages where there is one of the following restrictions on nasal sounds: (i) nasal consonants are nonphonemic (contextually predictable), or (ii) the opposition between nasal and oral vowels is neutralized after nasal consonants (in favor of nasal vowels).

Zusammenfassung: Vergleichende Daten aus verschiedenen Sprachfamilien zeigen, dass Nasalität zwischen dem silbeninitialen Konsonantenbündel und dem nachfolgenden Vokal transferiert werden kann. Ausgehend von den bisher beschriebenen Fällen wird anhand sino-tibetischer Daten eine neue Analyse vorgeschlagen. Der Transfer scheint in beide Richtungen zu gehen: vom konsonantischen Anlaut zum nachfolgenden Vokal für Tai-Kadai, Austroasiatisch, Sino-Tibetisch, Niger-Kongo (Kwa) und Indogermanisch (Keltisch), dagegen vom Vokal zum vorangehenden konsonantischen Anlaut im Sioux. Eine genauere Untersuchung der Bedingungen dieser Nasalitätsveränderungen fördert jedoch eine Asymmetrie zutage. In den meisten Fällen verläuft die Richtung des Nasalitätstransfers vom konsonantischen Anlaut zum folgenden Vokal. Alle Fälle eines regulären Transfers in die entgegengesetzte Richtung kommen aus Sprachen, in denen die Nasallaute einer der folgenden Beschränkungen unterliegen: (i) Die Nasalkonsonanten sind nicht phonemisch (d.h. sie sind kontextuell vorhersagbar) oder (ii) die phonologische Opposition zwischen nasalen und oralen Vokalen ist nach Nasalkonsonanten aufgehoben, zugunsten von nasalen Vokalen.

Résumé : Des données comparatives de plusieurs familles de langues montrent l'existence de transferts de nasalité entre un groupe de consonnes en position initiale de syllabe et la voyelle qui suit. Le passage en revue des exemples décrits à ce jour est complété par une nouvelle analyse de données sino-tibétaines. De prime abord, il semblerait que ce transfert puisse s'opérer dans les deux sens : de l'attaque consonantique vers la voyelle suivante – en tai-kadai, austroasiatique, sino-tibétain, niger-congo (kwa) et indo-européen (celtique) – et de la voyelle à la consonnes précédente en sioux. L'examen des conditions d'apparition de ces changements révèle néanmoins une asymétrie. Le cas de figure le plus courant est que le transfert de nasalité s'opère de l'attaque consonantique vers la voyelle qui suit ; les cas que nous avons pu trouver d'un changement régulier dans la direction opposée proviennent tous de langues dans lesquelles les sons nasals connaissent l'une des restrictions suivantes : soit les consonnes nasales n'ont pas valeur de phonèmes (c.-à-.d. que leur apparition est déterminée par le contexte), soit l'opposition entre voyelles orales et nasales est neutralisée après les consonnes nasales (en faveur de voyelles nasales).

Keywords: nasal onsets; nasal vowels; nasalization; consonant clusters; transphonologization; panchronic phonology; modeling of sound change


A widely attested diachronic change is the creation of nasal vowels from nasal codas, the latter disappearing in the process. Examples are found in many unrelated languages, for instance from Proto-Romance, which did not have nasal vowels, to modern Romance languages (Sampson 1999); in French, the masculine form of "good", /bɔ̃/ (spelt bon), alternates with the feminine /bɔn/ bonne. This phenomenon figures prominently in discussions of universals of nasalization (e.g. Hajek 1997).

The historical transfer of nasality between a consonant and a following vowel is actually attested, though it is less common. For instance, the transfer of nasality from an intervocalic consonant to a following vowel – and sometimes also to a preceding vowel – is observable as a synchronic alternation in Yal: /tʰiŋi/ "shellfish, Terebralia" has an alternant [tʰiɣ̃ĩ], and /waŋa/ "manner, way" an alternant [wãɣ̃ã] (Ozanne-Rivierre 1995: 54; see also Ozanne-Rivierre and Rivierre 1989); the nasal consonant turns into a nasalized spirant in the process.

The present article focuses specifically on the transfer of nasality between a consonantal onset and a vowel, raising the issue whether the evolution can go in both directions: from C to V, and from V to C. Section 1 presents cases in which the transfer is from C to V. Although superficially similar, the facts in Siouan (section 2) actually reveal a reverse development: the spread of nasality in Siouan is from V to C. It thus seems as if the transfer of nasality could take place in both directions. However, the discussion (section 3) points to a structural condition on the transfer of nasality between a complex consonantal onset and a vowel: in view of the data available to us, it appears that this transfer only takes place from the onset to the following vowel – except in languages where there is one of the following restrictions on nasal sounds: (i) nasal consonants are nonphonemic (contextually predictable), or (ii) the opposition between nasal and oral vowels is neutralized after nasal consonants.

1. Vowel nasalization from a consonantal onset

This section reviews cases of vowel nasalization from a consonantal onset; in view of these well-attested cases, a similar analysis is then proposed for a set of Sino-Tibetan comparative data.

1.1. The simplification of stop+nasal onsets in Kam-Sui (Tai-Kadai family, Southeast Asia)

The Lakkia language (a.k.a. Lajia) provides crucial insights into the origin of nasal vowels in the Kam-Sui subgroup of Tai-Kadai. The correspondences in Table 1 point to earlier initial clusters.

Table 1. Some correspondences between Sui (dialect: Sandong 三洞) and Lakkia, after Ferlus 1996: 255.1 Throughout the present article, bold type is used to draw attention to crucial examples.


Sandong Sui


Reconstructed initial cluster































Sandong Sui lost the stop part of the original cluster: the stop+nasal clusters *km-, *kn-, *tn- and *kɲ- merged with the preglottalized *ˀm-, *ˀn- and *ˀɲ- initials. The latter are preserved in Sui, e.g. /ˀma¹/ "vegetables", /ˀma³/ "flexible", both corresponding to a Proto-Kam-Sui * ˀm initial (Ferlus 1996: 251-252). Lakkia preserved the initial stop, while the nasal underwent lenition, nasalizing the following vowel in the process. Unexpectedly, the word "heavy" does not have a nasal vowel in Lakkia; such cases suggest that sporadic denasalizations took place after the creation of nasal vowels (Haudricourt 1967: 176).

Two Kam (a.k.a. Dong) dialects preserve forms that are very close to Lakkia – though without vowel nasalization – in the words "dog", "pig" and "flea": /kʰwa13/, /kʰu53/ and /kʰwat4/ respectively in Sanjiang Kam (Solnit 1988a: 234). Lakkia syllables with an initial velar stop stand in a regular relation of correspondence with nasal-initial syllables in Kam and in Southwestern Tai dialects (Solnit 1988b: 232-234; Edmondson and Yang Quan 1988; Ferlus 1996: 239; on similar facts in the Kra subgroup: Ostapirat 2000).

Data from Kam and Mulao reveal another type of change: distinctive nasality can spread to a preceding consonant. In Kam, *kʰm- > /ŋw-/; in Mulao, *kʰm > /ʰŋw/ (Ferlus 1996: 239-240). This is a striking structural parallel to the correspondences between two Austroasiatic languages, Laven and Nha Heun: the stop+nasal initial clusters of Laven correspond to nasal+medial in Nha Heun (Ferlus 1971). The evolution of medial nasals in Lakkia is summarized in Table 2.

Table 2. Evolution of medial nasals in Lakkia, after Ferlus 1996: 258.

Type of nasal consonant

Evolution in Lakkia




r (further changing to l or j)





The change found in Lakkia and Kam will be referred to below as lenition of medials, and the change in Sui as loss of cluster-initial consonant. Northern Sui dialects (Pandong 潘洞 and Yang'an 阳安) illustrate a possibility for the later evolution of glottal+nasal onsets: distinctive nasality is transferred onto the following vowel, and only the glottal remains, yielding [ʔṼ] or [h̰Ṽ]; the entire syllable is nasal, including the initial glottal sound (Haudricourt 1967: 176). The issue of the conditioning of the outcome of lenition (/h̰/ or /ʔ/) will be addressed in the general discussion, §3.3.2.

Download 392,16 Kb.
leave a comment
Date conversion02.10.2011
Size392,16 Kb.
TypeДокументы, Educational materials
Add document to your blog or website

страницы:   1   2   3   4   5   6   7   8   9
Be the first user to rate this..
Your rate:
Place this button on your site:

The database is protected by copyright ©exdat 2000-2014
При копировании материала укажите ссылку
send message