An Optimality Theoretic Account of Nasal Assimilation in English

Language, Information and Computation(PACLIC 11), 1996, 479-488 An optimality theoretic account of nasal assimilation in English* Seok-keun Kang Wonkwang University skkang@human.wonkwang.ac.kr. Abstract This paper reconsiders nasal assimilation in English within the framework of Optimality Theory and shows that the phenomena can be accounted for in a natural way in terms of some ranked violable constraints. In so doing, I also argue against an account based on structural interpretations of faithfulness, and propose that identity relation between input and output, i.e., correspondence, provides the best account. I also present further evidence that as pointed out by Lamontagne and Rice (1995), it is necessary to extend correspondence to the featural level. 1. Introduction The purpose of this paper is to reconsider nasal assimilation in English within the framework of Optimality Theory (hereafter OT; Prince & Smolensky 1993, McCarthy & Prince 1994). There has been in the literature much discussion on the above phenomena (e.g., Halle & Mohanan 1985, Borowsky 1986 among others). However, none of them provide a satisfactory account. In this paper, I will show that a purely constraint-based approach can account for nasal assimilation in a better way. In so doing, I will argue against an account based on structural interpretations of faithfulness, and propose that identity relation between input and output (i.e., correspondence, McCarthy & Prince (1994)) provides the best account. I will also show that as Lamontagne and Rice (1995) assert, correspondence should be extended to the featural level. The paper proceeds as follows. First, section 2 reviews previous analyses of English nasal assimilation. Second, section 3 presents the general principles of OT and discusses how the constraint-based approach can handle the phenomena under consideration. Finally, section 4 provides a brief summary of the paper. 2. Previous studies In English, the underlying nasals are /m/ and /n/. Unlike them, [o] does not exist as an independent segment. Rather it is derived from a sequence of nasal and velar obstruent by the so-called Nasal Assimilation. According to Halle and Mohanan (1985), the rule is formulated as (1). 479 (1) Nasal Assimilation (domain: stratum 2): In/� [o] / ___[-son, -cor, -lab] Sample derivations showing the operation of this rule follow: (2) /ink/� Auk/ (Nasal Assimilation) /long/ -4 /long/ (Nasal Assimilation)� /lou/ (/g/-deletion)' As Halle and Mohanan (1985) point out, however, the rule in (1) is blocked when the following syllable is stressed, as exemplified in (3b) in comparison with (3a); i.e., in (3a), In/ becomes [u], while it remains unchanged in (3b). 2 (3) a. co[n]gress co[n]cord sy[n]chrony i[n]cubate co[n]quer co[n]gruous co[n]crete b. co[n]gressional co[n]cordance sy[n]chronic i[n]clude co[n]cur co[n]gruity co[n]cretion Without giving a concrete account of the alternations in (3), they simply suggest that the determining factor appears to be the stress contour of the word; i.e., [u] does not occur before stressed vowels in English. However, this is not the case. Observe the examples in (4), where contrary to their assumption, [o] occurs before stressed as well as unstressed vowels. (4) li[n]guist� tra[n]quil�li[n]guistic�Hu[n]gary�Hu[n]garian tra[n]quility� Mo[n]gol�Mo[n]golian sa[n]guine� sa[n]guinity� bro[n]chi�bro[n]chitis�larkbalgophone lary[n]x …

Content truncated to 500 words