نشریه پژوهشهای زبانشناسی (Dec 2021)
Phonotactic Constraints on Tri-Syllabic Loanwords Containing Three-Consonant Sequences: An Optimality Account
Abstract
Abstract This paper examined the constraints on tri-syllabic loanwords in Persian, which contained three-consonant sequences in the underlying representation, within an optimality-theoretic account. Pulgram and Mich (1965:76) called a series of consonants as cluster occurring in the same syllable, and as sequence appearing in two consecutive syllables. Most of three-consonant sequences are the result of a complex onset placed word-medially in the second or third syllable and the consonants are put in the coda position of the preceding syllable. Having analyzed 9725 simple tri-syllabic words in Persian, a total of 36 words with three consonant sequences in the syllable boundaries were extracted. The results showed that they were all loanwords, except for three of them originating from Old Persian with no usage in contemporary Persian. In Persian, consonant clusters are avoided in onset position; consequently, three consonant sequences of a complex onset origin, when introduced into Persian, are adapted to fit the syllable structure of the target language and undergo re-syllabification. Re-syllabification occurs either with or without a vowel epenthesis and the processes all conspire to avoid a complex onset. Moreover, it is argued that in all these sequences, re-syllabification is subject to meeting the sonority at the syllable boundary and is explained by Gouskova’s (2001) proposal as the effect of the Syllable Contact Law (SCL), i.e., the preference for sonority to fall across a syllable boundary (Murray and Vennemann, 1983).Keywords: loanword, three-consonant sequences, Syllable Contact Law (SCL), re-syllabification. IntroductionIn Persian, any syllable in the surface realization contains an obligatory onset, an obligatory vowel as nucleus, and an optional coda consisting of one or two consonants, that is a structure of CV(C)(C) (Windfuhr, 1997: 138-139; Samareh, 1999: 108-110; Kambuziya, 2006: 149). Our study of simple tri-syllabic words with three-consonant sequences in Persian showed that they were all loanwords and behaved differently to adapt Persian syllable structure. Some three-consonant sequences resulted from a non-final CVCC and the following onset and they met the requirements of Persian syllable structure. However, most of the sequences resulted from an onset cluster placed word-medially. Onset clusters are not allowed in Persian, so when they are introduced into Persian, re-syllabification is inevitable. Re-syllabification includes movement of a consonant from one syllable to the adjacent syllable (Jam, 2015:22). To break up the onset cluster, some sequences are re-syllabified without vowel epenthesis; as in ‘institute’ [/Qns.ti.to], while in most loanwords, a vowel is inserted either between or at the edge of the two segments of the onset cluster as in ‘arthrose’ [/Ar.to.roz]. The different behaviors of these loanwords toward re-syllabification raised this question: What are the most plausible constraints governing re-syllabification in loanwords with three-consonant sequences of a complex onset origin? The hypothesis was as follows: Syllable Contact Law (SCL) explains re-syllabification either with or without vowel epenthesis in such sequences. Materials and MethodsThe research data included a set of tri-syllabic loanwords in Persian that contained three-consonant sequences in the underlying representation. To investigate the research hypothesis, 9725 simple tri-syllabic words were collected from Dehkhoda (2003) and Moshiri (2009) dictionaries, sorted in an Excel worksheet with all their etymological and phonological information, and then the three-consonant sequences were extracted. Except for 3 words, all the others were loanwords from English and French (a total of 36 words). The nativised pronunciations of the loanwords are mostly common and are based on the transcriptions provided in the aforementioned Persian dictionaries, as well as the linguistic intuition of the authors as Persian native speakers. The irrelevant details of segmental adaptations to fit Persian sound system do not concern the arguments. Discussion of Results & ConclusionsOur study of tri-syllabic loanwords containing three-consonant sequences showed that they were in an optimality account with re-syllabifications without epenthesis or with vowel epenthesis appearing to involve different processes, but conspiring to prevent onset clusters in Persian. The data provided enough support for Gouskova’s (2001) proposal that epenthesis stimulus is an onset cluster and that the Syllable Contact Law (SCL) plays a crucial role in re-syllabification, i.e., re-syllabification is to meet the sonority at the syllable boundaries. In this study, when an onset cluster displayed a segment sequence of falling sonority and consisted of a voiceless sibilant and a stop, re-syllabification happened without vowel epenthesis and the first consonant of the onset cluster moved to the coda poition of the preceding syllable. Therefore, the sonority fell at the syllable boundary without any vowel epnthesis. On the other hand, in some three-consonant sequences, the onset cluster displayed a segment sequence of rising sonority and consisted of a stop and a sonorant; hence, a vowel was inserted to break the onset cluster and the syllable contact costraints did not allow re-syllabification without vowel epenthesis and a rising sonority at the syllable boundary. In some sequences, however, the onset cluster consisted of three consonants, as a result of which a four-consonant sequence was formed. Thus, re-syllabification occurred both without and with vowel epenthesis so that not only SYLL-CON constraint was satisfied, but also unnecessary vowel insertion was avoided. Finally, the interaction of a set of markedness and faithfulness constraints, such as *COMPLEXONS, SYLL-CON, Dep-V, and *COMPLEXCOD, was argued to account for the sequences to be re-syllabified either with or without vowel epenthesis to avoid onset clusters.
Keywords
