Vowel Assimilation Research Articles

Phonology of Adur Niesu in Liangshan, Sichuan

This study describes the segmental and suprasegmental phonology of Adur Niesu, a Loloish (or Ngwi) language spoken mainly in Liangshan, Sichuan, southwest China. Phonemically, there are 41 consonants, 10 monophthongs and 1 diphthong in Adur Niesu. All Adur syllables are open. Its segmental changes mainly happen to the vowels, featuring high vowel fricativization, vowel lowering, vowel centralization, vowel assimilation and vowel fusion. It is common for Adur Niesu syllables to be reduced in continuous speech, with floating tones left. There are three main types of syllable reduction: complete reduction including the segment and tone, partial reduction with a floating tone left, and partial reduction with the initial consonant left. Adur Niesu employs tones as an important means for lexical contrast, namely, high-level tone 55, mid-level tone 33, and low-falling tone 21. There is also a sandhi tone 44. There are two types of tonal alternation: tone sandhi and tone change. Tone sandhi occurs at both word and phrasal levels, and is conditioned by the phonetic environment, while tone change occurs due to the morphosyntactic environment. Finally, some seeming tonal alternation is the result of a floating tone after syllable reduction.

DURASI ASIMILASI VOKAL TANWIN DALAM PEMBACAAN AL-QUR’AN SURAH AL- ISRA’ OLEH QARI’ INTERNASONAL

This study aims to analyze the duration of assimilation on tanwin vowel. The discussion studied in this study is the calculation of the duration of tanwin vowel assimilation. This study uses acoustic phonetic theory by applying qualitative descriptive methods through observation and data editing techniques. To analyze the data, there are four analytical techniques that are applied, namely the note-taking technique, the sorting technique, the digitization, and the sound segmentation. Meanwhile, for measuring the duration of the assimilation of sounds, this study used the praat program version 4.0.27. After the data analyzed, the results of the study showed that the duration of the reading of tanwin vowel assimilation by reciter 'DH sorted from the longest to the shortest, namely: For the duration of tanwin vowel assimilation read by reciter 'DH, from the longest duration lies to the shortest duration is seen in the word ʃaiɁĩn:faʂalnahu] has a duration of 3.171 md, the word [wa:ziratuw:wizra] has a duration of 2.448 md, the word [daraja:tiw:waakbaru] has a duration of 2.434 md, the word [xabi:ram:baʂi:ra] has a duration of 2.356 md, the word [faɖlam:min] has a duration of 2.186 md, the word [muɁminun:faɁula:Ɂika] has a duration of 2.029 md, the word [kita:bay:yalqahu] has a duration of 1.857 md, the word [maðmu:mam:madћu:ran] has a duration of 1.720 md, the word [kullan:numiddu] has a duration of 1.570 md, dan the word [ajraŋ:kabiran] has a duration of 1.287 md.

완도군 서부지역 자생해녀의 언어 연구

This paper analyzed the language of Native Haenyeo in the western regions of Wando-gun in grammatically, phonologically, lexically. In the Particle, if dialect type and standard language type coexist, there were locative case markers ‘-eseo’, Instrumental Case ‘-eulo’, Auxiliary particle ‘mada’, ‘buteo’. if there is only dialect type, there were Genitive case markers ‘ui’, locative case marker ‘ega’, Comparative case marker ‘cheoleom’, ‘mankeum’. In the Connective ending, if dialect type and standard language type coexist, there were ‘-eunikka’, ‘-eumyeon’. if there is only dialect type, there were ‘-eumyeonseo’, ‘-eulyeogo’, ‘-ji’. I have analyzed the language of Haenyeo phonologically, it was not different from the Wando dialect. In the vowel assimilation, in the case of two syllable stems, Bogildo Haenyeo appeared only ‘i+eo→yeo’ type. and Soando and Chuzado Haenyeo were coexisted with the ‘i+eo→e’ type and the ‘i+eo→yeo’ type. Compared to Wando dialect data, it can be estimated that the ‘i+eo→ yeo’ type is becoming more common in western Haenyeo. The w-glide formation was found in data from three regions of haenyeo. When the vowel of the stem syllable was ‘오’, the form in which the glide formation occurred and the form in which the glide ‘w’ was eliminated coexisted. and When the vowel of the stem syllable was ‘우’, the two forms will coexist. but the form in which the glide formation occurred was more common. The y-glide formation appeared in the data of haenyeo in three regions, Bogildo Haenyeo appeared only ‘i+eo→yeo’ type. and Soando and Chuzado Haenyeo were coexisted with the ‘i+eo→e’ type and the ‘i+eo→yeo’ type. Like the Wando dialect, the ‘i+eo→yeo’ type is likely to be more common after glide formation. I have analyzed the language of Haenyeo lexically. The western Haenyeo wad only called ‘Haenyeo’ and unlike Jeju Haenyeo, the high-skilled Haenyeo was called ‘Yeongja, Meoguli’ and the low-skilled Haenyeo was called ‘Banga’. Regarding the sea environment, in Bogildo island they works in the sand sea, in Soando island they works in the gravel sea, and in Chujado island they works in the ttenseom island. the vocabulary influenced by Jeju haenyeo was used, such as ‘meodeul’ or ‘neonji’. Regarding the sea space, Soando Haenyeo data showed ‘golaeyeo’ and ‘ungtong-gae’. Regarding working tools, the western haenyeo called the nets that contain seafood collected from the sea ‘mang-ali, mangsali, meongseoli, heongseoli’, and this is no different from Jeju Haenyeo. The harvesting tool was called ‘homu, kkalkku, kkakkuli’ and the tool that picks up abalone wad called the ‘pichang, binchang’. This seems to be influenced by Jeju Haenyeo. As a secondary tool, ‘dulumbag, dulbag’ are used for Haenyeo to rely or swim on the surface of the sea. this is different from Jeju Haenyeo. Soando Haenyeo called the line connecting the net and the anchor stone ‘ttasbae’.

Language Variation and Phoneme Distribution in Toba and Simalungun Languages

Variations of language that exist in society are systematic but not random, as what can be seen from Toba and Simalungun languages (TSL). This research was qualitative since it only collected and described TSL’s language variation and phoneme distribution. Data was taken from interview. The results show that vowel assimilation can be seen from TSL’s words horbo vs horbou, eme vs omei, and mate vs matei. Consonant assimilation exists in TSL’s combination, such as, baba vs babah, tubu vs tubuh, mangolu vs manggolu, manjolom vs manggolam, biang vs baliang, and unang vs ulang in which consonants /h/, /g/. /gg/, and /al/, and /l/ are found. Ellipsis appearing in /o/-/u/, /e/-/i/, /h/, /e/-/ui/, /b/-/l/ in Toba’s words such as horbo, eme, eta, marlange, and bagak can be found in Simalungun’s horbou, emei, heta, marlangui, and lagak. With reference to phoneme shift, several shifts in Toba’s consonant, such as, /rs/, /j/, and /e/ in words, such as marsuan, manjolam, and marlange have become /n/, /gg/, and /ui/ in words for example manuan, manggolam, and marlangui in Simalungun. Keywords: Language variation, assimilation, ellipsis, phoneme distribution, Toba and Simalungun languages

An Optimality Theoretic Analysis of the Mid Tone in Isekírì

The asymmetric behaviour of the mid tone has been reported in the Yoruboid literature and previous accounts of tone behaviour in Iṣẹkírì suggest that a low tone is always deleted in favour of others at word boundary. However, synchronic data have shown that this is not the case. This paper investigates tonal asymmetry and the non-participation of the mid tone in some phonological processes in Iṣẹkírì. The study adopts an Optimality Theory (OT) version that incorporates the grounding conditions to develop constraints of implicational statements into faithfulness constraints. Data are elicited from eight speakers of the language, recordings are with the aid of a digital audio device and the tool used for our acoustic measurement is Praat 6.0.30. The analysis reveals that tonal asymmetry in this language always deletes the Mid-tone at word boundary and not the Low tone, also noting that the Low tone may be deleted in some cases. The lack of participation of the M tone in some phonological processes such as vowel elision, glide formation and vowel assimilation reveals the unstable nature of this tone in Iṣẹkírì. This paper concludes that the M tone is in asymmetry to the H and L tones in Iṣẹkírì and demonstrates the need for more perceptual and acoustic studies of tonal behaviour in Iṣẹkírì.

Assimilation in the Shavshian dialect

We call Shavshian the speech of the inhabitants of historical Shavsheti. We note that the term Imerkhevian dialect was used to denote the dialect mentioned in the Georgian linguistic literature (N. Marri, Iv. Gigineishvili, V. Topuria, Iv. Kavtaradze, Sh. Putkaradze, T. Putkaradze ...), As much as it was believed that the Georgian language survived in only twenty villages in the Imerkhevi gorge. The term was first used by Niko Mari, others echoed his opinion. We introduce the term Shavshian dialect with the following consideration: A. Shavsheti (in today's sense) and Imerkhevi are parts of one historical Georgian country, Shavsheti. B. Georgian language is heard not only in the villages cultivated in Imerkhevi gorge, but also outside it: In Ustamisi, Chakvelta... C. Georgian vocabulary is embedded in the Turkish speech of the Shavshians. D. Georgian toponymy is still alive in the whole territory of Shavsheti; E. As part of Shavshian, as dialects, we also consider Machakhelian speech. For all the named reasons, we think it is better to use the term Shavshian dialect, in which, according to the available data, two dialects can be distinguished: Imerkhevian and Machakhelian. Shavshian is a pronounced dialect of the Georgian language. The authors discuss a single phonetic phenomenon based on the various materials obtained through expeditions arranged in the villages of Shavsheti - assimilation in Shavshian dialect. Cases of both vowel assimilation and consonant assimilation are discussed. There is a remarkable linguistic picture, namely: assimilation of vowel:

Regresyvinė balsių asimiliacija: nomenklatūrinė požymio kaita plungiškių požiūriu

    Straipsnyje nagrinėjama pagrindinio skiriamojo šiaurės žemaičių telšiškių ir kretingiškių patarmių požymio – regresyvinės balsių asimiliacijos – nomenklatūrinė kaita. Siekiama nustatyti, kiek Plùngės gyventojai suvokia regresyvinę balsių asimiliaciją kaip skiriamąją šiaurės žemaičių tarminę ypatybę, ir įvertinti šio požymio kaitą. Remiamasi dinamiškosios dialektologijos, perceptyviosios dialektologijos metodais.    Išanalizavus plungiškių informantų pastabas apie regresyvinę balsių asimiliaciją ir kitus tarminius šiaurės žemaičių patarmių požymius, nustatyta, kad balsių asimiliacija retai įvardijama šiaurės žemaičių patarmių skiriamuoju požymiu. Dažniausiai ši ypatybė pastebima, tačiau jai suteikiama nežymi diferencinė reikšmė. Svarbesniais, reikšmingesniais laikomi kiti fonetikos, morfologijos, leksikos skirtumai. Regresyvinė balsių asimiliacija taip pat siejama ir su netarminės kalbos įtaka.

Optimality Theoretical Approach in Resolving Vocalic Hiatus in Tiv Language

This paper investigates vocalic hiatus resolution strategies in Tiv language. Hiatus is a phonological phenomenon whereby two vowel sounds occur in adjacent syllables with no intervening consonant. It is a dispreferred configuration in many languages but when it inevitably occurs, it is resolved through some of these strategies: epenthesis, vowel assimilation, vowel elision, secondary articulation, vowel coalescence, glide formation and diphthongization. Though there are other hiatus resolution strategies in Tiv, for want of time and space, this paper examines the most common ones: vowel assimilation and vowel elision, as hiatus resolution/repair strategies in Tiv, through the lenses of the Optimality Theory. It adopts a descriptive and analytical research design. The intuitive phonological ability of the researcher and oral interview through purposive sampling were used for eliciting data. Data were presented and analysed in tables. The study establishes the occurrence of vowel hiatus in Tiv. It demonstrates that vocalic hiatus is dispreferred in Tiv as it is in many other languages of the world. The study has shown that Tiv utilises inter-alia vowel assimilation and vowel elision to resolve vowel hiatus. The resolution is meant to ease articulation of Tiv words. Well-formedness of vowel configurations in Tiv was found to be constraint based in which case optimality of a candidate that undergoes repairs through either vowel assimilation or elision is determined by the ranking of the universal constraints that interact in the selection of candidates.

A functional optimality-theoretic study on the aspect of development of vowel assimilation in the literary and colloquial reading of the Old Mandarin

A functional optimality-theoretic study on the aspect of development of vowel assimilation in the literary and colloquial reading of the Old Mandarin

The Historical Reconstruction of Greater Tzeltalan (Mayan) Vowel Assimilation and Vowel Raising Patterns

AbstractThis paper examines the phenomenon of vowel assimilation at a distance in di‐ and trisyllabic roots and stems in the Greater Tzeltalan languages of the Mayan language family. The study focuses on changes undergone in disyllabic and trisyllabic roots and stems, characterized by either anticipatory or perseverative vowel assimilation, and discusses the historical scenarios relevant to their reconstruction to the various relevant stages (Proto‐Greater Tzeltalan, Proto‐Ch’olan, Proto‐Tzeltalan), as well as likely cases of diffusion between Ch’olan into Tzeltalan. In addition, the paper also considers the question of vowel raising in Ch’olan, i.e. the so‐called *oo > u and *ee > i shifts, identified by Kaufman and Norman and reexamined by Brown and Wichmann. We posit a series of phonetic blocking conditions, and we argue that most cases can be accounted for as examples of suffix‐controlled vowel assimilation. It is also shown that the presence of certain typical suffixes with low or mid vowels can account for the etyma that did not experience the shifts.

A Critical Review of Some Contentious Issues in the Phonetics and Phonology of Ẹdo, Igbo, and Yoruba

Some contentious/controversial issues in the Phonetics/Phonologies of Ẹdo, Igbo, and Yoruba, (all members of the Kwa group of languages) are examined in this paper. This includes the appropriateness or otherwise of the use of the term “coalescence” in phonology based on available data in Ẹdo and Yoruba. The data from the Ẹdo language, and the reanalyzed Yoruba data do not seem to support the continued use of the term “coalescence” without any overt phonetic motivation as a tool for defining/describing what is actually vowel assimilation, vowel elision, and tone shift as a set of phonologically ordered processes. The second issue examined in this paper is the generally held view of the total assimilation of V₂ by V₃ across word boundary in a V₁CV₂ # V₃CV₄ collocation in Igbo. It is a known fact in languages that a more plausible (natural) assimilatory process involves a left to right movement not a right to left movement. It is argued that the so-called (total) vowel assimilation in Igbo V₁CV₂ # V₃CV₄ collocation is simply the case of the elision of V₂, and the re-association of the tone thereon with V₃, i. e., the first vowel of the second word. The autosegmental perspective is employed to elucidate this fact. In the third issue examined, it is argued that mutual exclusivity, used as the defining characteristic of languages that manifest vowel harmony, effectively excludes the Ẹdo and other Ẹdoid languages, in which there is free co-occurrence of vowels in any position of the word from being characterized as languages that manifest vowel harmony. The fourth issue examined is the manifestation of the downdrift and downstep phenomena. It is argued that the phenomena are language specific: non-phonemic in Ẹdo, involving a set of phonologically ordered rules whereas it is phonemic in Igbo. It is demonstrated that the only condition for obtaining a downstepped tone in Ẹdo is the presence of a H # L tone pattern across word boundary. The donwnstepped High tone in Igbo is phonemic irrespective of the nature (voice or voiceless) of the intervocalic consonants in VCV words.

A Description of Phonological Processes in ỌRO Compound Nouns

Compounding is a word-formation process. Though compounding is a morpho-syntactic phenomenon, its morpho-phonology interface cannot be gainsaid, as well-formed lexical items must be regulated by some phonological processes to comply with the phonotactic constraints of the language. This study describes the phonological processes that occur as a result of compounding in the Ọrọ language. The data for the study were elucidated from three selected native speakers of Ọrọ using the 1700 SIL Comparative African Wordlist, as well as a structured wordlist. The data were carefully coded and transcribed for analysis. The results of the analysis show cases of weakening, vowel deletion and vowel assimilation processes in Ọrọ compound nouns. These processes entail that Ọrọ phonotactics play a major role in defining a structurally-preferred lexical item in the Ọrọ language. It is therefore opined that compounding in Ọrọ is defined by the phonotactic constraints of the language. Analyzing the data within the moraic phonological framework clearly brought out the modifications that segments undergo, thereby making explicit, the alteration in the syllable internal structure of the compounds as maintained in this paper. This work will provide the basis for an aspect of phonological analysis in Ọrọ, as well as providing more insights into the structural requirements of the Ọrọ language.

Gradient and categorical assimilation of pretonic vowels in Brazilian Portuguese

This paper addresses the acoustic realisations of the pretonic vowels /e, o/ that have been previously reported to undergo regressive vowel harmony in Brazilian Portuguese. It examines how the height of pretonic /e, o/ is affected by the phonological and phonetic height of the adjacent stressed vowel in three dialects: Northeastern (Bahia), Northern (Amazonas) and Southern (Rio Grande do Sul). A pseudoword reading task was performed with two speakers each of the three different dialects. The findings suggest that there is some kind of low harmony, in that /e, o/ are realized with markedly higher F1 before the stressed low vowels /ɛ, a, ɔ/ than before the stressed non-low vowels /i, e, o, u/. This effect was found for all dialects, but appears to be categorical (and thus phonological) for the Northern and Northeastern speakers, while gradient for the Southern speakers, where it is likely due to phonetic V-to-V coarticulation. More importantly, no effect of height harmony was found in any of the dialects: pretonic /e, o/ were not produced significantly higher before a stressed /i, u/. In addition, Northern and Southern speakers showed V-to-V coarticulation for the non-high pretonic vowels, illustrating with Northern speakers that a categorical harmony process can co-occur together with a gradient vowel assimilation in the same dialect.

The geolect of Plungė in terms of regressive assimilation of vowels I, U

Based on dialectometric methods, the article discusses the geolect of Plungė in terms of regressive assimilation of vowels i, u. The study material consists of about 9 hours of audio recordings, 57 sentences, recited by nine presenters of younger, middle and older generations. 6 words were chosen in which regressive assimilation of vowels can take place, i. e., the words with vowels i, u in accented, unaccented and shifted accent positions. Quantitative analysis of the material (sentences read by the presenters) was performed with the tools of the computer program Gabmap. Pseudo maps of networks, reference points, cluster analysis, as well as differential dialectal features were analyzed. The analysis performed using dialectometry methods shows that differences in limb reduction, word stem, consonant softening become apparent, but in many cases regressive assimilation of vowels i, u becomes the main variable feature. The operation/inaction of the regressive assimilation of vowels i, u is greatly influenced by accent. When vowels are accented, presenters of all generations pronounce them without regressive vowel assimilation. When the vowel i is unaccented, it is assimilated, and the vowel u is spelled narrowly by only a third of the presenters. Dual behavior exists in cases where vowels receive a shifted accent. The pronunciation of both vowels is approximate. Maintaining the main distinguishing feature of the residents of Plungė from the dialect of the residents of Telšiai, although inconsistent, would allow predicting that the linguistic dialect peculiarity of this area could compete with the language code of Telšiai – based on the Samogitian regiolect – or whether the regiolect itself would be / become dual-core (more detailed research based on a multi-faceted research model is needed to confirm this statement). The effect of regressive assimilation in the Plungė dialect, in the geolectic zone in general, can be both a proof of resemblance to the northern Samogitian Telšiai residents and a sign of a decrease in the importance of assimilation as a distinctive feature of the dialects.

Dominance-as-markedness

This paper examines a formal consequence of the assumption that dominance is equivalent to markedness (Casali 2016): if dominant ATR values are marked and therefore specified, while recessive values are unmarked and unspecified, then no phonological process in a language with ATR dominance should require reference to the recessive value. This claim is examined in light of new data and analyses of ATR harmony and three other vowel assimilation patterns in Bari (Eastern Nilotic; BFA). I demonstrate that all four of these processes are analyzable without reference to the recessive value of ATR, supporting the characterization of dominance as markedness, and markedness as specification

Kleines Lautgesetz, große Wirkung

Zusammenfassung The unexpected i instead of expected *e in the first syllable of Latin wordforms such as nisi, nimis, nihil, mihi, tibi and sibi has until now been explained in various ways either from vowel assimilation of original *e to the i of the second syllable or from clitic weakening of the words. This article aims at giving a common mechanical explanation for all of these words by posing a new sound law according to which original *e in Latin becomes i in initial open syllables followed by a syllable which originally contained the vowel ẹ̄ as the result from monophthongisation of the diphthongs *ei̯, *oi̯or *ai̯(< *h₂ei̯) in second syllable position. For this purpose, the article reconstructs the history of the abovementioned explanations - especially the vowel assimilation theory which goes back to an early article by Sommer -, tries to falsify them by critically examining the evidence adduced for proving them, and eventually derives the new sound law from parts of the original evidence of the falsified explanations, and by making use of additional evidence.

Acoustic correlates of anticipatory and progressive [ATR] harmony processes in Ethiopian Komo

Komo, an endangered Koman language spoken in Western Ethiopia, features a system of vowel assimilation commonly referred to as ‘advanced tongue root’ (ATR) harmony. Prior accounts of Komo [ATR] harmony describe a typologically unique, bidirectional system with two distinct and productive processes: anticipatory [+ATR] spreads leftward to non-high vowels and progressive [−ATR] spreads rightward to high vowels. In this study, we investigated the acoustic correlates of the [ATR] feature in Komo using recordings from twelve native speakers collected in the field. Our aims were to describe the acoustic signature of the feature, evaluate acoustic evidence for the claim that both assimilatory processes indeed involve [ATR] spreading, and explore individual variability in the realization of the feature. The results of linear mixed effects models indicated that, in both processes, [+ATR] vowels featured lower F1 values, less periodicity, and a relatively pronounced first harmonic. The anticipatory process was also cued by duration differences while the progressive harmony featured a partial difference in spectral slope. As for individual strategies, random forest analysis revealed a great deal of variability in the relative importance of different correlates. While F1 dominated many of the acoustic profiles, some speakers relied primarily on voice quality or duration to signal the feature.

“Chamem a pulúicia!” harmonização vocálica progressiva no português da Madeira

Entre os principais fonólogos do português (por exemplo, Mateus e Andrade 2000), a Harmonização Vocálica Regressiva (HVR) tem sido assumida como o único tipo de assimilação vocálica no português europeu (PE). De facto, a Harmonização Vocálica Progressiva (HVP) é concebida como, se não impossível, pelo menos altamente excepcional, quase idiossincraticamente exclusiva de entradas lexicais específicas (ibid.). No entanto, em alguns dialetos do PE, a HVP revela-se muito frequente e produtiva. Este é o caso de alguns dialetos falados na ilha da Madeira (Segura da Cruz & Saramago 1999).

Acoustics of Tatar vowels: Articulation and vowel-to-vowel coarticulation

Volga Tatar is a Turkic language spoken by 5 million people in Central Russia for which instrumental acoustic descriptions are lacking. This study uses formant analysis of acoustic recordings from 27 native speakers of Volga Tatar to describe the vowels of Tatar and evaluate the accuracy of previous impressionistic phonetic descriptions. In addition to describing the acoustic vowel space of Tatar, the study uses carefully chosen target words to evaluate vowel-to-vowel coarticulation in height and backness among the Tatar vowels /i, ae, ɑ/. Examining vowel-to-vowel coarticulation in Tatar is of particular theoretical interest due to the presence of vowel harmony in the language. While the majority of native Tatar words are subject to backness, and possibly rounding, harmony, a large class of disharmonic lexical items, mostly from borrowings, provides insight into the coexistence of long-distance phonological vowel assimilation (vowel harmony) and long-distance phonetic vowel assimilation (coarticulation) in the same language. Previous research (Banerjee, Dutta, & S., 2017; Beddor & Yavuz, 1995) suggests that vowel harmony may suppress coarticulation proceeding in the same direction. However, the current results indicate that direction of coarticulation in Tatar is mediated primarily by other factors, such as target and trigger vowel identity.Volga Tatar is a Turkic language spoken by 5 million people in Central Russia for which instrumental acoustic descriptions are lacking. This study uses formant analysis of acoustic recordings from 27 native speakers of Volga Tatar to describe the vowels of Tatar and evaluate the accuracy of previous impressionistic phonetic descriptions. In addition to describing the acoustic vowel space of Tatar, the study uses carefully chosen target words to evaluate vowel-to-vowel coarticulation in height and backness among the Tatar vowels /i, ae, ɑ/. Examining vowel-to-vowel coarticulation in Tatar is of particular theoretical interest due to the presence of vowel harmony in the language. While the majority of native Tatar words are subject to backness, and possibly rounding, harmony, a large class of disharmonic lexical items, mostly from borrowings, provides insight into the coexistence of long-distance phonological vowel assimilation (vowel harmony) and long-distance phonetic vowel assimilation (coarticulation) in...

The morphology and phonology of metathesis in Amarasi

I describe and analyse data from Amarasi, a language with morphological consonant-vowel metathesis. Depending on the phonotactic structure of the stem to which it applies, metathesis is associated with a number of other phonological processes including: vowel deletion, consonant deletion and two kinds of vowel assimilation. By proposing that Amarasi has an obligatory CVCVC foot in which C-slots can be empty all these phonological processes can be derived from a single process of metathesis and one associated morphemically conditioned process. I consider analyses other than the rule-based one adopted in this paper and show that they cannot account for all the data in a consistent, plausible way.