Ioana Chitoran

Approaches to phonological complexity

The proposed volume draws on an interdisciplinary sketch of the phonetics-phonology interface in the light of complexity. Composed of several first-order contributions, it will consequently be a significant landmark at the time of the rise of several projects linking complexity and linguistics around the world.

From hiatus to diphthong: the evolution of vowel sequences in Romance

Romance languages show hiatus and diphthongal realisations of inherited iV sequences of rising sonority (e.g. ia). We study five Romance varieties with different degrees of contrast between the two realisation types: Romanian, with a diphthong–hiatus contrast, Spanish, with a weaker contrast, French, with no contrast (all diphthongs), and European and Brazilian Portuguese, with no contrast (all hiatus). We show that the different degrees of synchronic contrast are related to three independent factors : (i) a general articulatory tendency for [iV] hiatus to resolve to diphthongs, due to the relative stability of diphthongal articulations ; (ii) a structural ‘ attractor ’ effect of pre-existing [jV] diphthongs in a language, from different historical sources ; and (iii) prosodic lengthening effects which inhibit the shift from hiatus to diphthong, supported by phonetic studies of durational patterns across the five languages.

Towards standard measures of articulatory timing

Many studies have investigated the relative timing (or overlap or latency) between articulatory events to make inferences about coordination in speech production or about phonetic structure. A major problem is that these studies employ different measures of relative timing. One criterion for choosing a standard measure could be the variability exhibited across repetitions and talkers for a given set of gestures in a given phonological context. If a single measure appears to be the least variable across a variety of experiments, then it is a good index of the stable, phonetically relevant properties of coordination. In this study, 33 measures were compared using coefficient of variation. In four corpora of articulatory data on consonant clusters (EMMA and x‐ray microbeam), three articulatory events were defined: gesture onset, target, and release. Based on these events, coefficients of variation for the 33 measures were calculated across repetitions and speakers under a number of different linguistic condi...

Cross-linguistic differences in articulatory timing lag in consonant cluster perception

Cross-language research on consonant cluster production has shown that consonant clusters in different languages are produced with different degrees of articulatory timing lags. The present study examines perceptual sensitivity to these cross-linguistic timing differences in consonant clusters. Native German listeners were tested on an AXB similarity judgment test using stimuli including consonant clusters produced by German and Georgian speakers. (German consonant clusters are produced with relatively shorter lag between two consonants than Georgian ones.) Stimuli were /bla, gla, gna/ syllables recorded along with articulatory (EMA) data. Short lag German tokens and long lag Georgian tokens were selected as A and B, with Xs of varying degrees of lag chosen from either Georgian or German recordings. Results showed that German listeners are sensitive to the cross-linguistic differences in articulatory timing lag: when the timing lag of X was closer to A, participants were more likely to choose A. Moreover,...

Dholuo interdentals: Fricatives or affricates? Evidence from domain‐initial strengthening

The manner of articulation of 〈th,dh〉 in Kenya Dholuo is described controversially in the literature. This study uses acoustic data collected from three native speakers of the same dialect to test the hypothesis that 〈th,dh〉 exhibit positional variation, in accordance with domain‐initial strengthening. The prediction that 〈th,dh〉 are realized as affricates in strong position and fricatives in weak position was supported. All three speakers showed longer closure for 〈th〉 in pretonic than in non‐pretonic position (F=166.833, p IP‐medial word‐initial > IP‐initial without focus > IP‐initial with focus (F=116.513, p<0.01). IP‐initial 〈dh〉 and 〈d〉 had similar prevoicing duration (72.120; 72.434) suggesting that the interdental became more stoplike in this position. Furthermore, whereas it is known that vowels are longer before fricatives than before stops, the vowels before 〈th,dh〉 were not longer than those before the corresponding stops 〈t,d〉 suggesting 〈th,dh〉 are stop‐like. Using both qualitative and quantitative criteria, Dholuo interdentals are best described as affricates with differences in articulation in prosodically stronger versus weaker positions.

Prosodic effects on glide‐vowel sequences in three Romance languages

Glide‐vowel sequences occur in many Romance languages. In some they can vary in production, ranging from diphthongal pronunciation [ja,je] to hiatus [ia,ie]. According to native speakers’ impressionistic perceptions, Spanish and Romanian both exhibit this variation, but to different degrees. Spanish favors glide‐vowel sequences, while Romanian favors hiatus, occasionally resulting in different pronunciations of the same items: Spanish (b[j]ela, ind[j]ana), Romanian (b[i]ela, ind[i]ana). The third language, French, has glide‐vowel sequences consistently (b[j]elle). This study tests the effect of position in the word on the acoustic duration of the sequences. Shorter duration indicates diphthong production [jV], while longer duration, hiatus [iV]. Eleven speakers (4 Spanish, 4 Romanian, 3 French), were recorded. Spanish and Romanian showed a word position effect. Word‐initial sequences were significantly longer than word‐medial ones (p<0.001), consistent with native speakers more frequent description of hia...

An acoustic study of hiatus resolution in two Romance languages

Spanish and Romanian contrast vowel sequences [CiV] in hiatus and corresponding diphthongs [CjV], with some interspeaker variation. Both languages contain surface diphthongs derived historically by gliding, /CiV/ > [CjV]. They both show a strong tendency for blocking gliding word‐initially, supported by native speaker judgments: Sp. [miope], Rom. [miopu] short‐sighted; Sp. [italjana], Rom. [italjana] Italianf. Data from six speakers of each language confirmed this variation. The duration and F2 transition rate of the vocalic sequence were compared, in words containing [i.a] and [ja]. [ia] was significantly longer in hiatus, and had a significantly slower transition rate than in [ja]. However, the ranges of the hiatus and [ja] sets showed some overlap, suggesting that hiatus resolution is not a categorical phonological process. Instead, lexical items fall on a hiatus‐to‐[ja] continuum. Further comparison of ranges and standard deviations confirmed the correlation between this variation and word position. I...

Recoverability constraints on gestural overlap in Georgian stop sequences

Recent investigations of the temporal organization of articulatory gestures have found that consonants in a cluster exhibit less temporal overlap when in word‐onset position than when the cluster occurs word‐finally or spanning a word boundary. A possible reason for this is that substantial overlap of obstruent gestures in utterance‐initial position may threaten their perceptual recoverability. Recoverability considerations may also account for results showing that a front‐to‐back order of place of articulation in stop–stop sequences (labial–alveolar, alveolar–velar, labial–velar) allows more overlap than the opposite order. Presumably, the recoverability of C1 is hindered if the release of C1 has no acoustic manifestation due to the presence of a more anterior C2 being coproduced in time, hiding the C1 release. Data demonstrating both these constraints on gestural patterning have previously been drawn only from English, limiting the type and position of consonant sequences. The Georgian language offers a...