| old_uid | 4479 |
|---|
| title | The Pairwise Variability Index: rhythmic values from Italian Dialects and
Language Typological Implications |
|---|
| start_date | 2008/03/31 |
|---|
| schedule | 10h-12h |
|---|
| online | no |
|---|
| summary | In traditional views of rhythmic typology the rhythmic classification of a
language was considered a given ‘primitive’, an inherent property of the
language. In contrast, more recent studies see it as an emergent property,
a product of both the phonotatics of the language and phonetic processes
in production (Dauer 1987, Nespor 1990). Syllable complexity, possibly
vocalic and consonantal length distinctions, stress-dependent vowel
reduction and the propensity for phonetic vocalic and consonantal
reduction processes (‘schwa-isation’, weakening and elision, etc.) during
speech are considered contributory factors to the rhythm of an utterance,
and via this to the general rhythmic impression of a language. Critically,
rhythm not only becomes measurable (Ramus et al. 1999, Grabe-Low 2002,
Barry et al. 2003, Russo-Barry to appear) as a speech phenomenon rather
than an inherent language property, but it also necessarily becomes a
continuous rather than a categorical property.
A comparison of Standard Italian and Southern Italian dialects is carried
out using acoustic measures related to structural properties:
Vowel-interval and consonantal inter-vowel-interval durations are used to
obtain rhythmic measures based on a number of approach.
Italian is usually considered to be ‘syllable-timed’: it has a relatively
simple (CV-dominated) basic syllable structure, no phonological vowel
length opposition and no phonological vowel reduction. On the other hand,
it has a consonantal length distinction, and pronounced allophonic
tonic-vowel lengthening. The structural features found in the dialects
offer support for a divergence from the traditional assumption of
syllable-timing. Distributional observations and durational measurements
of tauto- and heterosyllabic VC sequences show that make a strictly
syllable-timed rhythmic structure untenable. Phonetic and phonological
evidence is presented to support the interpretation of Italian dialects as
a stress-timed language. Although much of the dialect observations place
the stress-timing evidence at the systemic rather than the realisational
level, the non-systemic, ‘performance’ evidence points in the same
direction.
Rhythm measures are calculated according to Ramus et al. 1999, Grabe-Low 2002, Barry et al. 2003, Russo-Barry to appear.
The Ramus measures are (i) the proportion of vowels in the interpause
streches ips (%V), (ii) the standard deviation of the Vowel duration in
the ips (delta V) and (iii) the standard deviation of the intervocalic
consonantal interval (delta C).
The Grabe and Low measures correspond in essence to the Ramus variability measures, but are calculated in pairwise steps through the ips rather than globally across the ips. They are therefore called ‘Pairwise Variability Indices’ (PVIs). The difference (i) between consecutive vowels and (ii) between consecutive intervocalic intervals are averaged over the ips, giving a vocalic and consonantal variability measure. In the case of the vowel intervals, the difference is related to the sum of the two vowels. This ‘normalisation’ is claimed to be necessary (and possible) for the vowel intervals in order to counteract shifts in tempo because vowels vary more than consonants with tempo, and there is never more than one vowel ina vowel interval.
We also present syllabically based PVI measures (PVI-Syll). It should be
pointed out that, while capturing sequential variation, the PVI fails to
maximise this possible advantage over the Ramus measures because vowel and consonant variation are calculated separately. The combined effect of vocalic and consonantal structure on an auditory rhythmic pattern is therefore not taken into consideration. PVI-Syll takes the consonant and vowel intervals together, reflecting the combined complexity of consonantal + vowel groupings in sequence within an interpause stretch.
In agreement with predictions derived from phonological observation, the
results of the above measures show a ‘rhythm plot’ in which the PVI-V and the PVI-Syll place the Italian dialect speakers nearer to the
‘stress-timed’ languages than traditional typology statements would lead
one to expect.
References
Asu, Eva Liina / Nolan, Francis (2005) “Estonian rhythm and the Pairwise
Variability Index”, Procedeengs Fonetik 2005, Department of Linguistics,
Götemborg University, 29-32.
http://www.ling.gu..se/Konferenser/fonetik/2005.
Barry, William John /Andreeva, Bistra / Russo, Michela / Dimitrova,
Snezhina / Kostadinova, Tania (2003) “Do Rhythm Measures Tell us Anything
about Language Type?”, Maria-Josep Solé / Daniel Recasens / Joaquin Romero
(eds.), Proceedings of the 15th Internaional Congress of Phonetic
Sciences, Barcelona, 3-9 August 2003, Barcelona: Causal Productions Pty
Ltd, 2693-2696.
Dauer, M. Rebecca (1987) “Phonetic and phonological components of language
rhythm”, 11th International Congress of Phonetic Sciences. Tallinn,
Estonia: U.S.S.R. / Academy of Science of the Estonian S.S.R., vol. 5,
447-450.
Grabe, Esther / Low, Ee Ling (2002) “Durational Variability in Speech and
the Rhythm Class Hypothesis”, in Carlos Gussenhoven / Natasha Warner
(eds.), Papers in Laboratory Phonology VII. The Hague: Mouton de Gruyter,
515-546.
Nespor, Marina (1990) “On the rhythm parameter in phonology”, Roca Iggy
M., (ed.), Logical Issues in Language Acquisition, Dordrecht:
Foris,157-175.
Ramus, Franck / Nespor, Marina / Mehler, Jacques (1999) “Correlates of
linguistic rhythm in the speech signal”, Cognition 73, 265-292.
Russo, Michela / Barry, William John (to appear) “Isochrony reconsidered.
Objectifying relations between Rhythm Measures and Speech Tempo”, Speech
Prosody 2008, 4th Conference on Speech Prosody, Campinas, May 6-9, 2008,
Brazil. |
|---|
| responsibles | Aroui |
|---|
| |