Probability Matching versus Probability Maximization in Morphophonology: The Case of Korean Noun Inflection (original) (raw)

1 Reanalysis and Regularization in Morphophonology : Korean Noun Inflection and Beyond

2018

The life cycle of phonological rules or processes is a concept that has been familiar to phonologists since the work of Kruszewski and Baudouin de Courtenay in the last decades of the 19th century. Coming into being with the phonologization (Hyman 1976) of mechanically determined phonetic variation and being as a result entirely general at the outset, phonological processes subsequently acquire morphosyntactic conditioning or come to apply in increasingly restricted morphological domains (Bermúdez-Otero 2007:504, 2015:382-384). Eventually, they may disappear from the grammar altogether, leaving traces of their former existence only in isolated alternations between lexically listed items; this is the case, for example, for the residue of Verner’s Law in English (was/were, lose/forlorn).

Base-Output Correspondence in Korean Nominal Inflection

Journal of East Asian Linguistics, 2006

This paper discusses various puzzles concerning the phonology of Korean nominal inflection. In particular, I investigate a range of vowel hiatus resolution phenomena that differ between nouns and verbs, the overapplication of consonant cluster simplification and laryngeal feature neutralization in nominal stems, and certain asymmetries between derived nouns and nonderived nouns. After presenting some problems with previous approaches, I offer an analysis of the phonology of Korean nominal inflection in terms of Base-Output Correspondence (BOC) Theory, along the lines of Kager (1999). I argue that a variety of unexpected properties of noun phonology in Korean can receive a unified account under the BOC approach. I also show that noun-specific alignment and noun-specific faithfulness constraints are inadequate to capture noun phonology in Korean. My arguments support the view that the fact that nouns have a Base (isolation form) is the key factor in explaining the unique properties of noun phonology (Kenstowicz's (1996) Base Identity, in particular). It is shown, however, that the BOC approach overcomes shortcomings with the Base Identity approach by making a necessary distinction between minor and fatal divergence from the Base. BASE-OUTPUT CORRESPONDENCE IN KOREAN NOMINAL INLFECTION 2 1. BASE-OUTPUT CORRESPONDENCE IN KOREAN NOMINAL INLFECTION 3 because of the requirement that nouns project a prosodic word. We can term this the Prosodic Approach which is represented by such works as Yongsung Lee (1999, 2001: Y. Lee, hereafter) and E. Kang (2000, 2001). Other researchers argue that nouns show peculiar behavior due to domain-specific faithfulness. We term this the Noun Faithfulness Approach which is represented by Smith (1997) and subsequent works. Still others argue that the complex behavior of nouns is traced to a morphologically related isolation form, the Base: the Base Identity Approach of Kenstowicz (1996). All these approaches seem to provide a plausible analysis for certain aspects of nouns in Korean. Crucially, however, the previous analyses do not adequately handle the full range of phenomena previously mentioned concerning the phonology of Korean nominal inflection. Furthermore, the data that each analysis covers do not overlap very much, and thus there has been no serious attempt to compare these competing analyses on the same ground. In this paper, I provide detailed arguments to show that on closer inspection, all the previous analyses have some problems in dealing with the puzzles concerning the noun-verb asymmetries in Korean. I propose a novel account for the phonology of Korean nouns by adopting Kager's (1999) Base-Output Correspondence (BOC) Theory. Specifically, I argue that the phonological complexity of nouns is tied to the morphological fact that nouns, not verbs, have a free-standing Base (cf. Kenstowicz 1996). Since nouns in Korean have a Base, they are significantly affected by the ranking of Base-output faithfulness constraints. Verbs in Korean, in contrast, have no Base, and thus they are not affected by Base-output faithfulness constraints. Consequently, the rankings of Base-output faithfulness constraints with respect to markedness constraints or inputoutput faithfulness constraints contribute to underapplication and overapplication of phonology BASE-OUTPUT CORRESPONDENCE IN KOREAN NOMINAL INLFECTION 4 in nouns, but not in verbs. Nonderived nouns behave like verbal stems because being unaffixed they too cannot be affected by Base-output (affixed form) faithfulness constraints. My proposal certainly inherits the insight of Kenstowicz's (1996) Base Identity approach concerning the special role of an isolation form in Korean phonology. However, it overcomes some remaining problems with the Base Identity approach. The original formulation of Base Identity requires complete identity between the output and the Base. Thus, it does not distinguish between minor and fatal deviation from the Base systematically (but see section 3.3.2 for discussions). The BOC analysis, on the other hand, evaluates the similarity between the output and the Base with respect to various relations defined in Correspondence Theory (McCarthy and Prince 1995, Kager 1999). Thus, under the BOC approach, the winning candidate may be quite different from the Base, yet be more faithful to the Base than the other losing candidates. In this paper, I show that BOC analysis resolves the puzzles that the previous Base Identity approach could not handle adequately-specifically, the cases in which some alternation from the Base is mandatory due to high ranked markedness constraints in Korean (e.g. intersonorant voicing). I also demonstrate that my proposal provides a unified account for a variety of noun-verb asymmetries in Korean, which have not been feasible under the previous analyses. BASE-OUTPUT CORRESPONDENCE IN KOREAN NOMINAL INLFECTION 6 2.1 Underapplication of Phonology in Nominal Inflection In Korean, vowel hiatus frequently arises when a vowel-final stem is followed by a vowel-initial suffix. The hiatus can be resolved by various processes such as glide formation, /ˆ/ deletion, and vowel coalescence (Y. Lee 1993). A noun-verb asymmetry is consistently observed in the hiatus resolution processes-particularly, in a stem-final position (cf. section 3.2.2 for hiatus resolution processes in nonderived contexts). Starting with glide formation, consider the verbal stems in (3) and the nominal stems in (4). When a stem ending with a high front vowel /i/ or a round vowel /u, o/ is followed by a vowel-initial suffix, the stem-final vowel is optionally turned into a glide: (3a). 2 Glide formation is obligatory if the output without glide formation yields two consecutive onsetless syllables: (3b). Crucially, however, glide formation is not applicable to a nominal stem-final vowel: (4)

The base of Korean noun paradigms: Evidence from tone

Korean Linguistics, 2014

This paper reports and analyzes the tonal patterns that emerge in South Kyengsang monosyllabic nouns that exhibit two well-known analogical changes in stem shape, one involving coronal obstruent codas and the other stems with an underlying cluster. By the first change, underlying and orthographic /nach/ ‘face’ inflects asnat̚,nach-ɨl(conservative) ornas-ɨl(innovative); and by the second underlying /talk/ ‘chicken’ inflects astak̚,talk-ɨl(conservative) ortak-ɨl(innovative). We find that many such nouns with a high-low tonal pattern change to high-high when inflected with the segmentally innovative stem. We propose that this tonal change supports the model of Korean noun paradigms proposed in Albright (2008) and Do (2013) in which the citation form serves as the base for the construction of the suffixed forms. If the base is a neutralization site, then learners select the alternant in which they have the greatest confidence of scoring a correct hit when undoing the neutralization.

Quantitative Research on Suffixes and Grammatical Patterns in Present-day Korean

Second Language Research, 2001

In this paper, I will present a quantitative analysis of the frequency and ranking of the following items in modem Korean: (l) prefinal and final endings (the latter are used as conclusive or non-conclusive morphemes), and (2) the grammatical patterns (including adverbial phrases that are connected with the sentential adverbs). In 1997, I conducted the first survey of agg lutinating suffixes, entitled "Quantitative Research on Verbal Morphemes in Modern Korean" (Lee 1997b). 001 FE127 ta t:J-: 6039 002 FE042 n L 5790 003 FE021 ko:il : 3843 004 FE093 nun ~ : 30~ 005 PFE015 ess ~ : 2454 006 FE371 e <>1 2341 etc. Since then a couple of others' studies on the frequency of suffixes have been reported based on their larger corpus but those of grammatical patterns not yet done by others but me as below in this paper. 001 Igr.-e(a) kata-<>1(o~) 7tt:J-741 003 1-ci(yo)-Al (k) 539 005 1-ci anhta(-cito anhta) 468-Al ~t:J-(-Al.£ ~t:J-) etc. 002 2-ko iss ta 004 1-e cwuta 006-e(a) ota (gr.=grade)-:il 9J. t:J-fi37-<>1 'ft:J-468-<>1(or) .2.t:J-428 The above findings are useful in the following ways: First, they are instructive in developing a rational basis for pedagogical syllabi and materials in teaching Korean as a foreign or a second language. Second, the findings can also be applied to analyzing the order of presentation of morphemes in elementary Korean materials for children in Korea. Third, the findings are relevant to machine translation because they provide infonnation on which Korean expressions are more preferable in translation.

Emergence of word prosody in (Seoul) Korean

Proceedings of International Conferences of Experimental Linguistics

It has been argued that Korean has recently developed an F0 distinction word-initially partially replacing the VOT distinction of the three stop categories, lax, aspirated, tense. This change has been characterized as tonogenesis, but since the contrast is not on all syllables, it seems to be more consistent with a pitch accent language than a tone language. We investigate the prosodic patterns of trisyllabic words to assess a) whether the VOT-to-F0 change is only word-initial or if it also occurs in other syllables, b) if there is evidence of word level prominence on one syllable supporting a pitch accent interpretation. The data from 10 Korean speakers yield conflicting evidence for both tonal and pitch accent prosodic systems.

A lexical account of inflectional suffixes in Korean

Journal of East Asian Linguistics, 1995

In this paper we give a reasonably comprehensive presentation of Korean inflectional suffixes, whose status has been somewhat controversial in the generative literature. We argue that nominal and verbal inflectional morphemes not only show phonological constituency with their hosts but also show evidence of lexical attachment, within the theory of Lexical Phonology and Morphology , , ). Consequently, the inflectional suffixes should combine with their hosts in a lexical formation component, separate from the syntax proper.

Assibilation or analogy?: Reconsideration of Korean noun stem-endings

Studies in Phonetics, Phonology, and Morphology, 2013

This paper discusses two previous approaches to the nominal stem-endings in Korean inflection, including loanwords: one is the assibilation approach, represented by H. Kim (2001), and the other is the analogy approach, represented by Albright (2002 et sequel) and Y. Kang (2003b). I contend that the assibilation approach is deficient in handling the underapplication to the non-nominal categories such as verb. More specifically, the assibilation approach is improper to explain why spirantization applies neither to derivative nouns nor to non-nominal items in entirety. By contrast, the analogy approach is able to overcome the difficulties involved with the assibilation position. What is crucial to the analogy approach is that the nominal bases end with t rather than with s. The evidence of tending bases is garnered from the base selection criteria, the disparity between tending and sending inputs in loanwords, and the alternations of stem-endings among t, s, ʧʰ, and ʧ. Unconventionally, I dare to contend that normative rules via orthography intervene as part of alternation extension, alongside semantic conditioning and token/type frequency.