Volume 13, Issue 2 (2022)
LRR 2022, 13(2): 33-69 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Fattahi M, Javaheri L. Phonological Opacity in Kermanshahi Persian: An Account Based on Optimality Theory-Candidate Chains. LRR 2022; 13 (2) :33-69
URL: http://lrr.modares.ac.ir/article-14-48027-en.html
URL: http://lrr.modares.ac.ir/article-14-48027-en.html
1- Assistant Professor of Linguistics, Department of English Language, Faculty of Literature and Humanities, Razi University, Kermanshah, Iran
2- PhD Candidate of Linguistics, Department of English Language, Faculty of Literature and Humanities, Razi University, Kermanshah, Iran ,l.javaheri@razi.ac.ir
2- PhD Candidate of Linguistics, Department of English Language, Faculty of Literature and Humanities, Razi University, Kermanshah, Iran ,
Abstract: (2828 Views)
The aim of this study is to analyze the role of phonological opacity in the phonetic representation of the prefix of progressive aspect in Kermanshahi Persian. In this dialect, when the progressive prefix /mi-/ is added to the verb stem, it changes in form in certain cases, depending on the nature of the first vowel of the verb. In Kermanshahi Persian, the progressive prefix is added to the left side of verbal root - which is, in turn, followed by verbal ending. This research is based on the Optimality Theory-Candidate Chains (McCarthy, 2007) - a version of Serial Optimality Theory. The data of this research have been collected through the conversations with the speakers of Kermanshahi Persian dialect in two age groups of over fifty- and under twenty-year-olds, relying – as well - on researchers' language intuition. The results of this study indicate that what causes the opacity in the phonetic representation of the vowel of progressive prefix in Kermanshahi Persian dialect is the occurrence of counter-bleeding interaction. The Serial Optimality Theory, is more successful in explaining this phenomenon than Parallel Optimality Theory.
1. Introduction
Kermanshahi Persian is a dialect which is spoken in the city of Kermanshah mainly in the west of Iran, and it presents certain linguistic differences with Standard Persian. One of the morpho-phonological differences is in the phonetic form of the progressive prefix /mi-/ when attached to certain verbal stems. Like what occurs in Standard Persian with the subjunctive prefix /be-/, the vowel of the prefix /mi/ undergoes vowel harmony if the following verbal stem contains a round vowel, and just as in /be-ro/ ("Go!") which surfaces as [bo-ro] on the phonetic level, /mi-/ in Kermanshahi Kurdish turns into [mo-] if there is a round element in the verbal stem, making forms like [mo.-ko.n-am] (meaning 'I do') out of /mi-kon-am/. The motivation of the such rule, however, is challenged by the phonetic form [mo.-xɑm] (I want), in which the vowel of the prefix is rounded, but there does not seem to be any round element as the motivation. This opaque application of the phonological process is analyzed in this article via the principles Optimality Theory. The main questions of the present study are as follows: (1) to what extent is Optimality Theory successful in explaining the opacity introduced by Kermanshahi Persian data? (2) What are the relevant active constraints? (3) In what permutation must the constraints be organized? (4) Which of the two versions of Optimality Theory, namely Parallel or Serial, is more successful in the explanation of the data in this study?
2. Literature Review
In order to prepare the grounds and to have a better insight into the present study, the review is based on the relevant works on phonological opacity and Optimality Theory which have been done inside and outside Iran.
2.1. International Studies
Bermúdez-Otero (1999, pp. 36-39) is among the first researchers who reported the challenge which the first versions of OT were faced with regarding the opacity phenomenon, and he asserts that the difficulty lies in OT neglecting the intermediate forms between the phonological and phonetic levels. This claim is also supported by McCarthy (2007, p.99), who reports the necessity of considering derivational path of the phonological processes. Urek (2013, p.1), too, reasons that Parallel OT fails to explain opacity since the Generator in this theory is designed to produce the candidates in one attempt. Anttila (2006, p. 893-901) classifies the suggested solutions to the problem of opacity in OT in three sets: "Input-Output Correspondence" theory which was first put forth by Benua (1995, pp. 77-82); "Sympathy Theory" suggested by MacCarthy (1999, 2000); and "Opacity Approach" put forward by Goldrick (2000, p.16). Alqahtani (2020, p. 22) is among the recent researchers who has investigated opacity as the result of counter-bleeding by analyzing examples from colloquial Persian. Also, Hauser and Hughto (2020, p. 29) claim that the general solution to opacity in OT could be found in what they call "contextual faithfulness constraints".
2.1. Iranian Studies
Many optimality-theoretic studies have been conducted in Iran on Persian and Kurdish data, among which one can mention those of Bijankhan (2006), Modarresi Ghavami (2011), Heidarpor Bidgoli (2012) in Persian; and Naqshbandi (2011), Fattahi (2012), Zahedi (2014), and Mohammadi (2014) in different dialects of Kurdish. As valuable as these studies are, none deals with the problem of opacity in Optimality Theory.
3. Methodology
The data in the present study have been collected in two ways: (1) through interviews with Kermanshahi Persian speakers of two age groups of above fifty and below twenty. The first group was chosen for their tendency towards using this dialect in their daily conversations, and the younger group was chosen as speakers who lean towards Standard Persian for its social prestige. This study, though, does not aim to investigate linguistic variations of these two age groups, but rather tries to identify the alternating forms in the verbal forms with progressive aspect. (2) through linguistic intuition of the researchers of the study, who are themselves, native speakers of Kermanshahi Persian. The collected data were then transcribed and the relevant alternating forms were identified. The last step was to identify the underlying phonological forms in order to decide on the nature and direction of the changes.
4. Discussion and Results
The following table illustrates the alternating forms of the progressive prefix in Kermanshahi Persian.
Table 1
The Alternative Forms in Progressive Prefix of Kermanshahi Persian Dialog
As can be seen from the data above, the alternating forms of the progressive prefix are [mi] and [mo], with the latter appearing where there is a round element in the verbal stem. The exception, though, is (11), in which the vowel of the prefix has been rounded without any apparent incentive. To have a better picture of what is going on in this form, we can compare the root with its likely historical form by synchronically comparing it with other Iranian languages and the forms in the history of Persian language. As evident even in the Persian orthography, there used to be the glide /w/ after /x/ sound in the verbal root of "to eat". If we take the /xw/ cluster as the onset of the root, then we can assume the line of the change as below (the verbal ending is removed for simplification):
/mi-xwɑ/ --(vowel rounding)à moxwɑ --(the omission of glide w)à[moxɑ]
If the glide omission, which is motivated by cluster simplification, was applied earlier than vowel rounding, the process of vowel rounding would be bled by it, leading to the incorrect phonetic form of *[mi-xɑ]; but since the actual order is in the opposite direction, the order of vowel rounding and glide omission is that of counter-bleeding, which causes opacity.
As was cited by many studies in the review section, parallel OT seems unable to choose correct opaque forms, the same is proved about Kermanshahi Persian data.
Tableau 1
Inability of Parallel Optimality Theory to explain Opacity Phenomenon
Optimality Theory-Candidate Chains, in addition to valuing the intermediate steps and derivational paths, enjoys a new form constraint that guarantees a fixed order of the application of the processes, namely Prec(A,B), in which constraint B is only violated if constraint A has been violated earlier. We can consider Prec(Ident(r),Max(r)) to penalize any candidate in which: (a) the round element is deleted without the process of vowel rounding having occurred earlier; (b) the round element is deleted and the vowel rounding occurs following that. With this constraint at play, we can consider the steps of derivation as depicted by the following tableaux.
Tableau 2
The First Step of Derivation
Tableau 3
The Second Step of Derivation
Tableau 4
The Third Step of Derivation and the End of Cycle (Convergence)
In Serial Optimality Theory, which OT-CC is a version of, the Generator is only allowed to produce candidates with one difference from the input. The winner chosen by the Evaluator will then be given to the same tableau, and there again, the Generator produces candidates with only one difference from the new input. This cycle will continue until the winner is the same form as the input – a step which is also called as convergence. This success in choosing the right form as the final winner is all thanks to the gradualness of Serial OT and the Prec constraint, which disqualifies the threatening competing forms Parallel OT was not able to shed out of the competition.
6. Conclusion
Optimality Theory-Candidate Chains proves to be more successful in the case of opacity observed in Kermanshahi Persian progressive prefix than the Parallel version of OT, where the process of choosing the final winner is not gradual and the intermediate forms are not taken into account due to the Generator's permission to produce candidates with any number of differences compared to the input. The gradual nature of OT-CC along with the Prec constraint leads to the choice of the right candidate as the winner. The following tableau, also known as Harmonic Improvement tableau, shows how the winners of the later rounds incur less threatening violations.
Tableau 5
The Harmonic Improvement in [mo.xɑm]
In the tableau above, the candidates are the winners of each step of derivation, with each being more harmonic – incurring fewer violations to the higher-ranking constraints – than the previous candidate.
1. Introduction
Kermanshahi Persian is a dialect which is spoken in the city of Kermanshah mainly in the west of Iran, and it presents certain linguistic differences with Standard Persian. One of the morpho-phonological differences is in the phonetic form of the progressive prefix /mi-/ when attached to certain verbal stems. Like what occurs in Standard Persian with the subjunctive prefix /be-/, the vowel of the prefix /mi/ undergoes vowel harmony if the following verbal stem contains a round vowel, and just as in /be-ro/ ("Go!") which surfaces as [bo-ro] on the phonetic level, /mi-/ in Kermanshahi Kurdish turns into [mo-] if there is a round element in the verbal stem, making forms like [mo.-ko.n-am] (meaning 'I do') out of /mi-kon-am/. The motivation of the such rule, however, is challenged by the phonetic form [mo.-xɑm] (I want), in which the vowel of the prefix is rounded, but there does not seem to be any round element as the motivation. This opaque application of the phonological process is analyzed in this article via the principles Optimality Theory. The main questions of the present study are as follows: (1) to what extent is Optimality Theory successful in explaining the opacity introduced by Kermanshahi Persian data? (2) What are the relevant active constraints? (3) In what permutation must the constraints be organized? (4) Which of the two versions of Optimality Theory, namely Parallel or Serial, is more successful in the explanation of the data in this study?
2. Literature Review
In order to prepare the grounds and to have a better insight into the present study, the review is based on the relevant works on phonological opacity and Optimality Theory which have been done inside and outside Iran.
2.1. International Studies
Bermúdez-Otero (1999, pp. 36-39) is among the first researchers who reported the challenge which the first versions of OT were faced with regarding the opacity phenomenon, and he asserts that the difficulty lies in OT neglecting the intermediate forms between the phonological and phonetic levels. This claim is also supported by McCarthy (2007, p.99), who reports the necessity of considering derivational path of the phonological processes. Urek (2013, p.1), too, reasons that Parallel OT fails to explain opacity since the Generator in this theory is designed to produce the candidates in one attempt. Anttila (2006, p. 893-901) classifies the suggested solutions to the problem of opacity in OT in three sets: "Input-Output Correspondence" theory which was first put forth by Benua (1995, pp. 77-82); "Sympathy Theory" suggested by MacCarthy (1999, 2000); and "Opacity Approach" put forward by Goldrick (2000, p.16). Alqahtani (2020, p. 22) is among the recent researchers who has investigated opacity as the result of counter-bleeding by analyzing examples from colloquial Persian. Also, Hauser and Hughto (2020, p. 29) claim that the general solution to opacity in OT could be found in what they call "contextual faithfulness constraints".
2.1. Iranian Studies
Many optimality-theoretic studies have been conducted in Iran on Persian and Kurdish data, among which one can mention those of Bijankhan (2006), Modarresi Ghavami (2011), Heidarpor Bidgoli (2012) in Persian; and Naqshbandi (2011), Fattahi (2012), Zahedi (2014), and Mohammadi (2014) in different dialects of Kurdish. As valuable as these studies are, none deals with the problem of opacity in Optimality Theory.
3. Methodology
The data in the present study have been collected in two ways: (1) through interviews with Kermanshahi Persian speakers of two age groups of above fifty and below twenty. The first group was chosen for their tendency towards using this dialect in their daily conversations, and the younger group was chosen as speakers who lean towards Standard Persian for its social prestige. This study, though, does not aim to investigate linguistic variations of these two age groups, but rather tries to identify the alternating forms in the verbal forms with progressive aspect. (2) through linguistic intuition of the researchers of the study, who are themselves, native speakers of Kermanshahi Persian. The collected data were then transcribed and the relevant alternating forms were identified. The last step was to identify the underlying phonological forms in order to decide on the nature and direction of the changes.
4. Discussion and Results
The following table illustrates the alternating forms of the progressive prefix in Kermanshahi Persian.
Table 1
The Alternative Forms in Progressive Prefix of Kermanshahi Persian Dialog
| (1) | mi.-bi.n-am prog. – see – 1S "I see." |
(2) | mi.-xɑ.n-am prog. – read – 1S "I read." |
|||
| (3) | mi.-bu.s-am prog. – kiss – 1S "I kiss." |
(4) | mi.-ka.n.-am prog. – dig – 1S "I dig." |
|||
| (5) | mi.-ʃu.r-am prog.- wash – 1S "I wash." |
(6) | mi.-ke.ʃ-am prog.-pull-1S "I pull." |
|||
| (7) | mo.- xor.-am prog.-eat-1S "I eat." |
(8) | mi.- sɑ.z –am prog.-make-1S "I make." |
|||
| (9) | mo.- ko.ʃ- i prog.-kill-2S "You kill." |
(10) | mo.- ko.n- am prog.-do-1S "I do." |
|||
| (11) | mo.- xɑ- m prog.-want-1S "I want." |
(12) | mo.- wo.r -am prog.-cut-1S "I cut." |
|||
| * prog.: progressive – 1S: first person singular – 2S: second person singular | ||||||
/mi-xwɑ/ --(vowel rounding)à moxwɑ --(the omission of glide w)à[moxɑ]
If the glide omission, which is motivated by cluster simplification, was applied earlier than vowel rounding, the process of vowel rounding would be bled by it, leading to the incorrect phonetic form of *[mi-xɑ]; but since the actual order is in the opposite direction, the order of vowel rounding and glide omission is that of counter-bleeding, which causes opacity.
As was cited by many studies in the review section, parallel OT seems unable to choose correct opaque forms, the same is proved about Kermanshahi Persian data.
Tableau 1
Inability of Parallel Optimality Theory to explain Opacity Phenomenon
| Ident (r) | Aligh-left(r) | mi-xwɑ-m |
| * | a. mixwɑm | |
| * | b. moxwɑm | |
| Mc. mixɑm | ||
| * | d. moxɑm | |
| Notes: symbol M shows an incorrect winning candidate. - Aligh-left(r) constraint penalizes any candidate with no agreement in rounding - Ident (r) penalizes any candidate with a changed value in the feature [round] | ||
Optimality Theory-Candidate Chains, in addition to valuing the intermediate steps and derivational paths, enjoys a new form constraint that guarantees a fixed order of the application of the processes, namely Prec(A,B), in which constraint B is only violated if constraint A has been violated earlier. We can consider Prec(Ident(r),Max(r)) to penalize any candidate in which: (a) the round element is deleted without the process of vowel rounding having occurred earlier; (b) the round element is deleted and the vowel rounding occurs following that. With this constraint at play, we can consider the steps of derivation as depicted by the following tableaux.
Tableau 2
The First Step of Derivation
| Ident (r) | Max (r) | No-Codaprg | Prec(Id,Max) | Complex-Syll | Align-l(r) | mi-xwɑ-m |
| * | * | Fa. mox.wɑm | ||||
| * | *! | b. mi.xɑm | ||||
| * | *! | c. mix.wɑm | ||||
| * | *! | d. mox.wɑm | ||||
| Notes: Complex-Syll penalizes any candidate in which there is consonant cluster in a syllable - No-Codaprog penalizes any form with a coda in the same syllable as the progressive prefix- Max (r) penalizes any candidate in which a round segment has been deleted (compared with the input) | ||||||
The Second Step of Derivation
| Ident (r) | Max (r) | No-Codaprog | Prec(Id,Max) | Complex-Syll | Align-l(r) | mox.wɑm |
| * | Fa. mo.xɑm | |||||
| *! | b. mo.xwɑm | |||||
| *! | c. mox.wɑm | |||||
| * | * | *! | d. mix.wɑm |
Tableau 4
The Third Step of Derivation and the End of Cycle (Convergence)
| Ident (r) | Max (r) | No-Codaprog | Prec(Id,Max) | Complex-Syll | Align-l(r) | mo.xɑm |
| Fa. mo.xɑm | ||||||
| *! | b. mo.xwɑm | |||||
| *! | c. mox.wɑm | |||||
| *! | d. mi.xɑm |
In Serial Optimality Theory, which OT-CC is a version of, the Generator is only allowed to produce candidates with one difference from the input. The winner chosen by the Evaluator will then be given to the same tableau, and there again, the Generator produces candidates with only one difference from the new input. This cycle will continue until the winner is the same form as the input – a step which is also called as convergence. This success in choosing the right form as the final winner is all thanks to the gradualness of Serial OT and the Prec constraint, which disqualifies the threatening competing forms Parallel OT was not able to shed out of the competition.
6. Conclusion
Optimality Theory-Candidate Chains proves to be more successful in the case of opacity observed in Kermanshahi Persian progressive prefix than the Parallel version of OT, where the process of choosing the final winner is not gradual and the intermediate forms are not taken into account due to the Generator's permission to produce candidates with any number of differences compared to the input. The gradual nature of OT-CC along with the Prec constraint leads to the choice of the right candidate as the winner. The following tableau, also known as Harmonic Improvement tableau, shows how the winners of the later rounds incur less threatening violations.
Tableau 5
The Harmonic Improvement in [mo.xɑm]
| Ident (r) | Max (r) | No-Codaprog | Prec(Id,Max) | Complex-Syll | Align-l(r) | |
| * | a. mix.wɑm | |||||
| * | * | b. mox.wɑm | ||||
| * | c. mo.xɑm |
In the tableau above, the candidates are the winners of each step of derivation, with each being more harmonic – incurring fewer violations to the higher-ranking constraints – than the previous candidate.
Keywords: Serial Optimality Theory, Optimality Theory-Candidate Chains, phonological opacity, Kermanshahi Persian
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |