Earticle

This paper analyzes the typology of assimilation (voicing assimilation and place assimilation), focusing on the 'grandfather effect' (McCarthy 2002, 2003a) observed in the case where the assimilation to marked structures is not allowed but the assimilation to unmarked structures is allowed. It recognizes that Hall (2007) accurately points out a flaw in McCarthy (2002, 2003a) where the typological difference in assimilation cannot be explained. This paper, however, shows that contrary to Hall's argument, the flaw has not been caused by the separation of markedness constraints, and proposes that IDENT(F) in McCarthy's analysis should be replaced with MAX(F) constraints. This proposal means that IDENT(voice) should be replaced with MAX(voice) in voicing assimilation and IDENT(place) should be replaced with the universally fixed constraint ranking MAX(dors), MAX(__cor) » MAX(lab) » MAX(cor), MAX(__lab|dors), which originates from Jun (1995), to retain the separation of markedness constraints. Further, this paper shows that the separation strategy of IDENT(F) also accounts for the grandfather effect cases in assimilation.

This paper investigates the sprouting type of sluicing to see whether sprouting always fails to nullify the (strong) island violation effects. Contrary to Chung, Ladusaw and McCloskey's (1995, 2006) predictions, sluicing with a deeply embedded implicit antecedent becomes possible when an implicit correlate of a remnant is activated by the preceding context. Furthermore, sprouting is sensitive to weak islands which do not block movement of a wh-phrase in its fully fledged counterparts as discussed in Romero (1998). This means that the extension of the form-chain algorithm to sprouting cases is neither necessary nor sufficient for analysis of sluicing wih implicit antecedents.

The purpose of this paper is to suggest the objectives, achievement standards, teaching materials, and teaching-learning methods in the English National Curriculum based on multicultural education. In order to achieve this purpose, the theoretical grounds for multicultural education are reviewed, and the 2007 English National Curriculum is analysed. It was argued that the objectives, teaching materials, and teaching-learning methods are to revised considering multicultural education, because helping students learn to understand multiple perspectives on events and phenomena is critical to citizenship education, and learners should be engaged in multicultural education. Furthermore, it was analysed that the components of the 2007 National Curriculum Revised are necessary but not sufficient for a comprehensive multicultural education in primary and secondary schools. Thus, in this paper, the directions of revising the English national curriculum considering multicultural education are suggested systematically and efficiently.

Modern Korean has four adnominal forms: '-(u)l', '-(u)n', '-nun' and '-ten'. The aim of this paper is to clarify the functional meanings of four adnominal forms. The conclusions we have reached are as follows: (1) '-(u)l', '-(u)n', '-nun' are the adnominal forms which consist of one morpheme. Whereas '-ten' is adnominal forms which consists of two morphemes: '-te-' and '-(u)n'. (2) '-te-' is the pre-final ending and belongs to a category of mode. It has the function that shows 'evidentiality' related to events or states of propositions what the speaker realizes. '-te-' is marked form of the evidential mode. When it does not appear, the epistemic mode which has the function that shows speaker's judgements about the events or states of propositions is shown. So the epistemic mode is unmarked in modern Korean. (3) '-te-' has a function to switch a scene. Therefore, the scene which a verb or an adjective combined with '-te-' shows is different from a scene at the time of the utterance by all means. (4) We consider the adnominal endings to prescribe domains. “The domains” that we speak here are the place where a speaker recognizes description contents (propositions). '-nun' has the function to show the domain handling realized events surely, whereas '-(u)n' has the function to show the domain to treat events that were already realized. '-nun' and '-(u)n' are similar to show 'realis'. However '-(u)l' has the function to treat events that are not yet realized and to do not understand whether they were realized in, so '-(u)l' is different from '-nun' and '-(u)n' in its point to show 'irrealis'.

Considered in the context of phonetic variability, inter-gestural overlap in the articulation of C1C2 sequences is gradient (e.g., more overlap and less overlap). EMMA is used to collect direct kinematic data from eight Seoul-Korean speakers to examine if gestural overlap differs depending on various factors. Four consonantal sequences (/tk/, /tp/ ,/pk/, /pt)/ are recorded. Focusing on a case study of Korean, we investigate order effect (e.g., more overlap in a front-to-back sequence compared to the reverse order) and phonological context effect on overlap. We also examine linguistic (word boundaries) and paralinguistic effects (speech rate and repetition). The results of the current study show that Korean speakers have internalized different phonological status, producing more overlap in the assimilation context than in the non-assimilation context. For both linguistic and paralinguistic factors, more overlap is obtained in the within-word condition and in the fast rate condition with inter-speaker variability.

It has been widely believed in the field of Korean linguistics that Korean Language has a tense system as in European languages. I argue that Korean Language does not have a tense as a grammatical category. Under this hypothesis, the morphosyntactic status of the Korean temporal elements is characterized as the following. 1)'-nun-' is not a morpheme as a syntactic unit. If we assume that '-nun-' is a single morpheme that is positioned in a node on a syntactic structure, we lead to a totally ad hoc account of the realizations of the allomorphs. 2)'-te-' is not a morpheme as a syntactic unit, on similar grounds. 3)'-essess-' is not a combination of two syntactic units, but one morpheme marking a subcategory of aspect. 4)'-ko iss-' and '-e iss-' are not single syntactic units, but elements that belong to the different phrases, building into a complex sentence structure. 5)The adnominal clause marker '-un' and '-ul' are morphophonological realizations of '-ess-nun' and '-keyss-nun', respectively. 6)'-ess-' is not a tense, but an aspect element.

Transparent free relatives (TFRs) are distinguished from standard free relatives (SFRs) in that they contain a ‘small clause’ consisting of a wh-gap and a predicate phrase, and more importantly, the predicate phrase behaves as the syntactic and semantic ‘nucleus’ of the free relative. Previous analyses have either treated TFRs as a unique construction that necessitates such mechanisms as parenthetical placement or grafts, or assimilated them to SFRs, relying on abstract/empty head elements and a vague semantic relation holding between the gap and the predicate phrase. By contrast, the present paper proposes that, while TFRs share basic constructional properties such as filler-gap dependency and exocentricity with SFRs, their syntactic head values and semantic interpretations should be determined by specific lexical and constructional constraints that pertain to the elements of TFRs. Within the framework of HPSG, the paper shows that the transparency effect of TRFs can be handled by feature inheritance from the nucleus predicate phrase, together with a constructional constraint that deals with the exocentric property of TFRs.

The purpose of this paper is to understand the category of adjective properly by investigating the interrelationship between semantic and grammatical properties. The following respects–semantic difference according to position, meaning types and grammatical properties, distinguishing indicators between participial adjectives and verb participles, and syntactic and semantic characteristics according to adjectival complementation types–are dealt with in this study. As a result, it should be noted that we need to recognize the importance of the interaction between grammar, meaning, and lexical idiosyncrasy. The discussion of grammaticality without considering the lexical and semantic properties of adjectives has some limitations for proper understanding and use.

This study presents that Korean Initial Tensification should be considered from perceptual perspectives. To be more precise, the strengthening resulting from tensification facilitates the word recognition. Along this line, it is proposed that word-initial position as well as initial syllable is an eligible strong position, rectifying the stances held by Beckman (1998) and Barnes (2002). Basically, we claim that the left-edge tensification is a compromise between two contradictory demands: From the paradigmatic dimension, it happens at the price of neutralization of plain and tense consonants. Meanwhile, under the syntagmatic dimension, it contributes to enlarging the concerned acoustic modulation, that is, perceptual fluctuation between preceding consonants and the following vowels. A priority is given to the increase in the syntagmatic modulation at the expense of neutralization of paradigmatic dimension. On the other hand, the option for tensification rather than aspiration derives from an Optimality perspective. Considering the evidence of phonetic measurements of VOT and F0, it turns out that the discrepancy between Korean aspirated and plain consonants is greater than in the case of tense and plain consonants. Under the control of faithfulness constraints, the tense consonants are expected to emerge.

Traditional accounts of the Old English breaking suggest that breaking is epenthesis of a protective back vowel and that the segments before which this epenthesis takes place are specified as [+back]. This paper aims to inquire into the nature of the epenthesis and explain the causes of breaking and subsequent sound changes within the framework of the Site-Articulator model and the Optimality Theory. I argue that the triggering element of the epenthesis is the velarization of the coda consonants, which requires the insertion of a floating labial constriction. It will be shown that the seemingly complicated and confusing sound changes are due to the interactions of several universal constraints. I also make a critical review of a couple of previous studies on the Old English breaking and try to show how my analysis overcomes some of the empirical and theoretical problems they potentially face.

This article examines the variation displayed by believe-type verbs between the complement clause involving Raising-to-object (RtO) and the complement clause without raising in Korean. Through a corpus study on the distribution of the complement clauses of the verb sayngkakha- 'think', we demonstrate that the choice between the two types of complement clauses is strongly affected by the contrastiveness of the proposition expressed by those clauses. Going one step further, we also provide a possible explanation for why the two types of complement clauses are distributed the way they are. We propose that the pattern of complement choice observed in the corpus data reflects the speaker's attempt to balance between processing cost and discourse function. In particular, it is shown that the balance between processing cost and discourse function can correctly predict the gradient pattern of complement choice, while offering a unified explanation for the effect of contrastiveness on complement choice and the 'characteristic property' condition on RtO.

This paper examines how the global use of English is changing the culture of personal naming. In Hong Kong, the majority of people use English first names along with their Chinese surnames. An analysis of job application documents and personal interviews with 19 speakers uncovered many interesting aspects about Hong Kong's naming culture, such as duplicate names and name change. Deeper investigation revealed that a number of linguistic and non-linguistic factors, such as Hong Kong’s unusual political status in China, the fact that Cantonese, the local dialect, has inferior linguistic status than Mandarin, its history as a British colony, and the current internationalized way of life - have converged to foster the use of English and English first names.

We will provide a clausal predicate analysis of Multiple Nominative Constructions (MNCs) in Korean, under the spirit of phrase structure grammar approaches. MNCs in Korean have two interesting characteristics: there is no limit on the number of nominative NPs in the construction, and the nominative NPs in MNCs are not directly related to each other. The purpose of this paper is twofold. Firstly, we will provide new syntactic mechanisms that capture the two characteristics above. In doing so, we will compare MNCs with multiple accusative constructions (MACs) in order to better understand their syntactic and semantic characteristics. Secondly, regarding the grammatical function of the nominative NPs of MNCs, we will show that the initial NP does not have a special status, e.g., as a "Large/Major Subject." It will be proven to be a regular subject of the clausal predicate on its right, just as each of the other NPs are the subject of the (clausal) predicate on its right.

In this article, I claim that there are two types in Korean double nominative constructions in which the leftmost nominal expression exhibits an GEN/NOM alternation. In one type, the first nominative nominal is a grammatical subject, while the second one is a complement. In the other, both nominals are "subjects," as a specifier of an IP or an AgrP. I also claim that there is a correlation between an inherent nature of the second nominative nominal and a type of a Korean DNC. When the nominal denotes a property, the sentence is of the first type, and when it denotes a kinship or an alienable possession, the sentence is of the second type. Finally, when it denotes a part, the sentence is structurally ambiguous.

Scrambling in Korean and Japanese has long been treated as an instance of movement, in an attempt to make all the languages share the same configuration in a deep level. In the paper, I claim that English and Korean/Japanese use different mode of syntactic combination instead of having the same configuration. While predicates select arguments in English, arguments select predicates in Korean and Japanese. The information of case markers helps them to find an appropriate predicate. The argument centered combination allow free word order in those languages. The order between arguments does not affect selection of a predicate. The subject-object order and the object-subject order selects the same ranges of predicates; transitive or ditransitive predicates. Long scrambling uses the information of the complementizer ko. It fills the semantic value(s) of a direct object, an indirect object, or both of them when its clause misses them, using the semantic value(s) of argument(s) in the left.

Flemming and Johnson (2007) have proposed that there is a fundamental distinction between a mid central vowel [ə] and a high reduced vowel [ɨ] in that [ə] occurs in unstressed word-final position while [ɨ] occurs elsewhere. Their study was the first instrumental study on the distribution of English reduced vowels but seemed to be insufficient to generalize them, since their work focused on the two reduced vowels - [ə] and [ɨ] - using minimal-pair data such as Rosa's vs. roses. Therefore, the present study revisits their study to explore the precise nature and typology of American English reduced vowels. For this, ten native speakers of English produced five types of test materials in which the factors of phonological position, morphology, and orthography were considered. First two formant frequencies were measured and their scatterplots were drawn for analyses. The results found three reduced vowels [ə], [ɨ], and [ɪ] and showed that [ɨ] occurs in the word-internal position while [ə] occurs elsewhere. The reduced vowel [ɪ] can appear together with [ə] in the word-initial position. It was also revealed that English orthography was a significant factor but morphology was not in the typology of English reduced vowels.