An Optimality Theoretic Account of Place and Voice Assimilation in Bushehri Dialects

1000 | P a g e c o u n c i l f o r I n n o v a t i v e R e s e a r c h A p r i l 2 0 1 6 w w w . c i r w o r l d . c o m An Optimality Theoretic Account of Place and Voice Assimilation in Bushehri Dialects Aliyeh Kord-e Zafaranlu Kambuziya*, Pegah Aryaei 2 , Sahar Bahrami Khorshid 3 1 Associate Professor of General Linguistics, Tarbiat Modares University, Tehran, Iran. Email: akord@modares.ac.ir 2 Graduate student in General linguistics, Tarbiat Modares University,Tehran, Iran. email: pegah.aryaee@gmail.com 3 Assistant Professor of Linguistics, Tarbiat Modares University, Tehran, Iran. email: sahbahrami@modares.ac.ir


INTRODUCTION
The purpose of the present study is to provide an overview aboat one of the common phonological processes in Bushehri dialects: assimilation, and to review the theoretical implications of this process. Bushehr province is located in south Iran. Northward, it is bounded to Khuzestan and Kohkiluye-Boyer-Ahmad, Eastward to Fars Province, Southward and Westward to the Persian Gulf. Bushehr have 10 counties.

Syllable Structure
The Bushehri dialects syllable may have three components: a peak, an onset and a coda. The nature of these components is as follow: "th p ak, o n cl , alway th mo t p om n nt l m nt of th yllabl . t m t b compo d of a vowel, either long or short. The marginal elements, on the other hand, contain only consonants, the onset always con t ng of a ngl con onant and th coda con t ng of z o, on , o two con onant ." Put differently, the following formula can be utilized to describe the syllable structure in Bushehri: C1VCo-2.
a.CV: a n th wo d: /ko/ "mo nta n" b.CVC: a n th wo d: /d m/ "ta l" c. CVCC: a n th wo d of /vaxt/ "t m "

THEORETICAL FRAMEWORK
The theoretical framework used in this study, namely Optimality Theory (OT), came into existence in early 1990s, mainly focusing on contemporary phonological studies (Prince and Smolensky, 1993;McCarthy and Prince, 1999;Mac Carthy, 2001). Th th o y wa d v lop d a a pon to a ''conc pt al c at th c nt of phonolog cal tho ght" (P nc and Smolensky, 1993) concerning the role of output constraints. It was also inspired by the concepts of neural networks, as shown by the significance of ideas such as optimization, parallel evaluation, competition, and conflicting constraints which are also present in the framework of OT. Optimality Theory is often considered as a development of generative grammar and the successor of the harmonic grammar developed in 1990s (Legendre et al., 1990(Legendre et al., , 2001Prince and Smolensky, 1993;Smolensky and Legendre, 2006, a.o.). In a typical OT analysis, the phonological constraints are ranked and violable by the phonetic forms of their underlying Representations in a tableau. This is a constraint-based competition system among a possibly infinite set of candidates (at least two) that are actually the potential surface forms of the input. The candidates minimally violate the constraints and the one that incurs the least serious violations in terms of hierarchies of constraints, wins. Hence, logically, the violations of higher-ranked constraints are most serious and these violations first oust the concerned candidate from the competition and so on. An OT-style tableau, which uses harmony maximization as the criterion for optimality where the weights are in the top row and the rightmost column, provides the harmony values for the candidates (Kar, 2009). This framework can effectively analyze various aspects of phonological issues in a language in a comprehensive and systematic manner.An outline of th "cla c" OT look l k a follow: np t → GEN → cand dat t → EVAL (con t a nt ) → Opt mal O tp t

METHODOLOGY
The method of this research is analytical-descriptive. In order to conduct the study, ten native speakers, selected from different age groups (from 20 to 60 year-old ones) and different educational levels were interviewed. The corpus primarily contains free conversation and life stories.
Here, McCarthy ' s (2008) combination tableau was adapted . The combination tableau illustrates the ranking between constraints, as well as violation marks. In the tableau, each losing (L) candidate is compared to the winning (W) candidate in regards to each constraint. (W) Denotes that the constraint in question prefers the winner rather than the losing candidate. This is why the winner satisfies the constraint but the losing candidate does not, as specified by the violation mark (*).Whereas the (L) denotes that the given constraint prefers the losing candidate rather than the winner (W). Bakovic (2007:335) assumes that the process of assimilation can be generally partitioned into two major types, local and long-distance. The cases of local assimilation take place rigorously between adjacent segments, such as between consonant segments within a consonant cluster. On the other hand long-distance assimilations occur between segments, whether consonants or vowels which are not adjacent such as consonants across a vowel. Crystal (2008:40) illustrates that several classifications of assimilation can be found. One of the classifications is due to whether the segment shifting is the consequence of the effect of a neighboring sound or of one that is not adjacent. The first type which is local assimilation is the common one. Crystal calls it contiguous or contact assimilation, and calls the other type non-contiguous or distance Assimilation. Lass (1984:171) give an example of long distance assimilation in which the sound /n/ changes into /m/: open/ əpən əpəm/. Here, the sound /n/ is changed into the /m/ because of the influence of the sound /p/. Crystal (2008:40) gives an example for long distance assimilation in which again the sound /n/ changes into /m/: turn up trumps /tε: np tmps/ tε: mp trmps/. The sound /n/ in turn has been changed into /m/ as a result of an influence of /p/ in up. He further argues that long distance assimilation occurs in languages having vowel harmony, in which a vowel in one part of a word may influence other vowels to be articulated similarly, despite the fact that there are other sounds separating the assimilated vowels.

Assimilation
There is another form of categorizing assimilation according to the direction of assimilation. Carr (1999:16) states that in regressive assimilation the first of two sounds undergoes assimilation to a coming sound. He also considers this type as the most common type of assimilation. According to Trask (1996: 26) regressive assimilation is a phenomenon in which a phonetic attribute extends to a preceding segment, which he also calls anticipatory co-articulation. Crystal (2008: 40) argues that there are three possibilities in what concerns the direction of assimilation. To him, the first one is regressive or anticipatory as explained above, the second type is progressive assimilation in which a sound changes as a result of the effect of the preceding sound. The occurrence of this type of assimilation in English is very few and it is not common. The th d typ coal c nt o c p ocal' assimilation in which there is a shared effect or mixture of the sounds on one another. For example: ten bikes /ten baɪks /→ ɪks/.
The direction of assimilation here is from left to right (regressive or anticipatory assimilation) in which the sound /n/ in ten has been assimilated to the sound /m/ as a result of the influence of the /b/ in the word bike.
Another taxonomy of assimilation is related to the influence of the sound features on the process of assimilation. Lass (1984:173) states that the process of assimilation greatly influences almost all the sound segments. The above classifications of assimilation can be according to two major parameters, place and voice. In place assimilation the assimilating segment spreads the feature of place onto the assimilated sound. If the assimilating segment is bi-labial, as a result of the assimilation process the assimilated sound will copy the place feature bi-labial from the assimilating segment as: ten bikes /ten baɪks m baɪks/.
Here the sound /n/ in ten is an alveolar sound, while the sound /b/ in bike is bi-labial, that is why the sound /n/ has been assimilated to a bi-labial sound which is /m/.
The other type is voice assimilation, Carr (1999: 16) explains that voice assimilation is a common kind of the process in which the assimilated segment takes voice feature from the assimilating sound. Katamba (1989:81) explains that the plural ma k " " n Engl h a f n xampl of voice assimilation; for instance: The last point about the classification of assimilation will be about whether the assimilated sounds are totally assimilated or partially. Crystal exemplifies the case by stating that in an instance like: ten bikes, /ten baɪks m baɪks].
The sound /n/ is assimilated to the sound /b/ only in the place feature not in both features of place and voice that is why the assimilation is partial not total. But in the example: ten mice, /ten maIs/ → [t mmal ].
The sound /n/ in the word ten is changed to the sound [m]. Here the assimilated sound is totally similar to the assimilating sound that is why the assimilation is total.

Place assimilation (acutness and graveness features)
Hyman (1975 :31) a g that: " nce the earliest phonetic studies, segments have been classifiied according to their articulatory properties. In consonants, for example one asks where a sound is made (place of articulation), how it is made (manner of articulation), and what the state of glottis is (voiced, unvoiced, ect.). in vowel, one asks which part of the tongue is raised (front, back, central), how much it is raised (high, mid, low), and whether the lips are rounded. While this is the most common and oldest way of classifying sounds, it is now possible with technological advances to group sounds according to their acoustic properties. That is phonetic features such as the one distinguishing In other words, segment can be similar (or dissimilar) either in the way they are made or in the way they sound two aspect which a lat d." A number of phonological properties cannot be accounted without considering the acoustic properties of the sounds in qustion. A simple case is seen in the following data from some varietis of Bushehri such as Deylami, Dashtestani, Liravi and Shuli: While a front vowel might be accpected to be backed before a back (velar) consonant, the changes of [u] to [i] before mentioned consonants is not so easily explained. It would appear that these consonants, which function together in this fronting process, have some phonetic feature in common, and yet articulatorily they are made at the same extreme in the oral cavity.
The reason is that the consonants such as {j, c, t, d, n, r, l, z, s, ʤ, ʧ} share an acoustic property which the other consonants like {b, p, m, x, γ} do not ha . S nc lab al and v la con onant a mad at th p f of th o al cav ty (two at th back of th mo th {x, γ}, om at th lab al {b, p, m}) produces a concentration of energy in the lower frequencies of the sound spectrum. Since alveolar/ dental and palatal sounds cut the oral cavity in two parts, they do not create a large oral cavity, but rather two smaller cavities. Consequently, they have in common a consentration of energy in the upper frequencies of the sound spectrum. This acoustic distinction is directly incorporated into the feature system proposed by Jakobson et al. Labial and velar consonants are said to share the property of graveness (low tonality) and alveolars and palatals (also palato-alveor and all coronals in case of Bushehri varities) share the property of acuteness (high tonality).
Back vowels like labial and velar consonants, are made at the periphery of the oral cavity, since the tongue is raised in the back of the mouth; front vowels, like dental/alveolar and palatal consonants, are made in non-peripheral (or medial) part of the oral cavity, since the tongue is raised in the center of the mouth. Consequently, both consonants and vowels differ in this acoustic property of graveness/ acuteness, as follows:

Gerave
Acute Labial consonants Dental/alveolar consonants Velar consonants Palatal consonants Back vowels Front vowels The data in table (2) shows that a grave high vowel becomes an acute high vowel before an acute consonant. This formulation reveals that the process in question is phonetically motivated: back vowels change to agree in acuteness with the consonants after them. This phenomenon involve cases of assimilation by which segment acquires the feature(s) of It is assume that assimilation between adjacent segments is driven by the family of agreement constraints (Lombardi 1996ab, 1999; see also Beckman 1998, Butska 1998, Bakovic´2000) defined in (1)

Voice Assimilation
As we explained before, one type of assimilation is voice assimilation. In bushehri dialects like Dashtestani, Ddeilami,Gami,ect. When voicless palatal [c] placed before a voiced consonant, it takes voice feature from the voiced consonant. Table ( 4) shows this type of assimilation.

CONCLUSION
One of the taxonomy of assimilation is related to the influence of the sound features on the process of assimilation. This type of classification of assimilation can be according to two major parameters, place and voice. In Bushehri dialects there is a place assimilation in acuteness feature. Hyman (1975 :31) argues that both consonants and vowels differ in this  . Consequently, both consonants and vowels differ in this acoustic property of graveness/ acuteness, as follows: grave (labial and velar consonants, back vowels) and acute (dental/ alveolar, palatal consonants, front vowels).