Search

Formosan languages, spoken by the indigenous peoples of Taiwan, have unique roles in the reconstruction of Proto-Austronesian Languages. This paper presents a real-world Formosan language speech dataset, including 144 h of news footage for 16 Formosan languages, and uses self-supervised models to obtain and analyze their speech representations. Among the news footage, 13 h of the validated speech data of Formosan languages are selected, and a language classifier, based on XLSR-53, is trained to classify the 16 Formosan languages with an accuracy of 86%. We extracted and analyzed the speech vector representations learned from the model and compared them with 152 manually coded linguistic typological features. The comparison shows that the speech vectors reflect Formosan languages' phonological and morphological aspects. Furthermore, the speech vectors and linguistic features are used to construct a linguistic phylogeny, and the resulting genealogical grouping corresponds with previous literature. These results suggest that we can investigate the current real-world language usages through the speech model, and the dataset opens a window to look into the Formosan languages in vivo. [ABSTRACT FROM AUTHOR]

Since the late 1980s, Taiwan has moved away from Mandarin-only language policy in favour of greater recognition of local Taiwanese languages as part of a greater localisation movement. While continuing to implement language policies aimed at promoting local Taiwanese languages, in December 2018, Taiwan announced intent to implement a bilingual language policy to incorporate English by 2030, titled Bilingual Nation. Applying Ball's (1993) policy as discourse framework, this paper endeavours to investigate the Bilingual Nation policy planning and rhetoric so as to shed light on Taiwan's official conceptualisation of national identity. The case study finds that English is the primary focus of Bilingual Nation policy documents and is relegated to the discursive arenas of internationalisation and global economic competition. English is rarely discussed in tandem with any other specific language or language category, with little consideration for the interaction of English and Taiwan's official national languages. In sum, Bilingual Nation acts as an affirmation of the current status quo in Taiwan's official conception of national identity: muddled, with the assumption of Mandarin at its centre. Questions remain for Bilingual Nation and its ability to push beyond a symbolic state effort for internationalisation or unproductive labours for international status. [ABSTRACT FROM AUTHOR]

Mainland Chinese director Wang Quan'an's Apart Together , which won the Silver Bear Award for Best Screenplay in 2010, tackles the issue of cross-Strait relations by telling the story of Kuomintang veteran Liu Yansheng's return to mainland China after nearly 40 years of separation from his wife and son. Shanghainese is the main language of the film, a dialect that is used to suggest a local attitude towards the national issue. While earlier Chinese films on similar themes often emphasise the cultural and emotional ties between Taiwan and China, in Wang's film Liu is characterised as an unwelcome Taiwanese guest, an intruder in his wife's Chinese family. This essay argues that Wang's Apart Together contests the People's Republic of China's official discourse of cross-Strait reunification by demonstrating the cultural and identity divisions between the Taiwanese character and his Chinese family. Wang provides an alternative perspective on the 'Taiwan issue', showing that ordinary people's experience of cross-Strait reunion might be painful and problematic. [ABSTRACT FROM AUTHOR]

This article revisits the debate which occurred in the early 1930s regarding 'Taiwanese Language Writing' to argue that these discussions represented an extended interrogation of the linguistic ecology of the island: how written scripts and spoken sounds functioned in Taiwan to make meaning in relation to a larger environmental, animistic, and social whole. Set against the backdrop of this debate, the 1932 work Elegant Words by the famed intellectual Lien Heng stands out as an important text. Lien Heng's vision of Minnan as a historically grounded yet broadly cosmopolitan language was a particularly enabling expression of Taiwanese consciousness, which presented an island that could face many directions at once, absorbing the textual signs of a deep Sinographic past while remaining vitally open to a complex and increasingly integrated world. Lien Heng's cultural criticism presaged many of the debates about language, history, and identity that would take place in subsequent decades in Taiwan, and as such his work continues to have resonance today for thinking critically about Taiwan's place in the modern world. [ABSTRACT FROM AUTHOR]

This paper proposes an analysis of VP fronting in the VOS Austronesian language Seediq in the theory of Labeling. In the same vein as Massam (Nat Lang Linguist Theory 19:153–197, 2001), I propose that verb-initial word order results at least in part from fronting the verb and object together in the VP when the object is indefinite and nonspecific, but the object vacates the VP when it is specific or definite in order to value case. Following this, either verbal head movement or remnant VP fronting takes place. However, I depart from Massam's approach when it comes to the motivation for VP fronting. First, I argue that VP movement does not target [Spec, TP] but rather moves no higher than the edge of the clause-medial phase. Consequently, the motivation cannot be satisfaction of the EPP, as Massam claims. Secondly, I show that whether verb-initial word order is obtained through verbal head movement or phrasal VP fronting is a consequence of Labeling. In particular, I go beyond Massam's approach and account for the fact that when the VP contains more than one internal argument, one of them must move to the case-checking position and the other must be pied-piped with the verb to the phase edge. Otherwise, it will not be possible to label the original VP. [ABSTRACT FROM AUTHOR]

We show that some elements in Mandarin that have been considered as pre‐verbal adverbial modifiers are actually the main verbs of the sentences. Three sets of phenomena are discussed: the degree gou 'enough' construction, the "Adv/Adv‐di" contrast, and two types of "manner wh‐adverbs". These phenomena make Mandarin a typologically interesting case, as it stands in contrast to languages such as English on the one hand, where adverbs are adjuncts, and to Formosan languages on the other, where manner modifiers realize as the main verb of the sentence. We also discuss the possible parameterization in grammar that is responsible for this distinction. We suggest that grammatical theory provides an option to languages, in the form of a parameter, which licenses complementation of an un‐subcategorized expression YP to a head X, when X denotes a manner (or more generally, an event modifier) of YP. [ABSTRACT FROM AUTHOR]

Taiwanese, a variety of Southern Min, is a dialect spoken in Taiwan with a pervasive tone sandhi system. To speak Taiwanese fluently and appropriately, one has to frequently apply Taiwanese tone sandhi. While it is common knowledge that tone sandhi is pervasive in spoken Taiwanese, there are no studies reporting on the rate of Taiwanese tone sandhi in spoken corpora in any genre. This paper attempts to determine the average tone sandhi rates of L1 Taiwanese dominant speakers and L1 Taiwanese L2 Mandarin bilinguals by analyzing a spoken corpus. Our study shows that the two groups do not reveal any significant differences in their tone sandhi application rate. On average, out of every 100 syllables, approximately 67 syllables have undergone tone sandhi. Regarding the application rates of individual tone sandhi rules, all speakers show a consistency as to which rules are most common and which are not. Consequently, one implication of our study is that L2 Mandarin fluency does not impact L1 Taiwanese tone sandhi frequency. Other implications help make clear the specific nature of the Taiwanese tone sandhi system. [ABSTRACT FROM AUTHOR]

Copyright of Taiwan Journal of Linguistics is the property of Graduate Institute of Linguistics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The Formosan language Tsou embraces three serial verb constructions where verbs bearing nonagreeing voice marking are strung together. While the combination of a nonactor voice first verb and an actor voice second verb appears to be an exception to the widespread tendency that components within a serial stricture have the same value for grammatical categories typically associated with verbs, this study argues that the nonconcordant voice marking actually reflects the syntacticosemantic conditions for integrating verbs as well as the event descriptions in their denotations into a joint predication, and adjustment in the pattern of argument realization jointly attests to the presence of these conditions. In Tsou, a nonconcordant verb series that introduces thematically peripheral, nonsubcategorized entities into a monoclausal structure is only well formed if the participating verbs do not clash in event properties crucial to the Vendler-Dowty style of lexical aspectual classes. On the one hand, constraints on the use of voice marking and combinatorial possibilities of verbs show that serialized verbs share common specifications for telicity and duration. On the other, changes in the morphosyntactic expression of arguments illustrate the possibility to shift aspectual characteristics in the verb's denotation so as to bring event descriptions within the entire verb series to a unified integration. As a whole, the Tsou data make evident the presence of event integration within verb serialization, even though the nonconcordant voice marking appears to suggest otherwise. [ABSTRACT FROM AUTHOR]

Under a generative framework, this paper investigates the properties of the AVonly restriction on complementation and how it comes about across many Formosan languages. It is found that the AV-only complements occur as a nonfinite defective vP without any formal feature and that they are tenselss and subjectless--temporal markers must attach to the matrix verb; the embedded object, if any, is required to move to the matrix subject position to get case. The AV-only restriction is attested not only in restructuring constructions but also in non-restructuring ones. In both types, the matrix verbs are highly restricted: in the first type, the matrix verb is restricted to a restructuring verb and in the second type, the matrix verb is limited to a verb that projects an external argument. In restructuring contexts, the matrix external argument, if there is one, is inert and creates no intervention effects for the shifted embedded object; in non-restructuring contexts, the matrix external argument blocks the raising of the embedded object, thereby preventing the embedded verb from occurring in the non-AV transitive form. The AV-only restriction is due to multiple factors, most important of which is a locality condition. [ABSTRACT FROM AUTHOR]

This paper is a study of Information Structure (IS) in Northern Amis and its interaction with the voice system. These two domains are distinct, but interacting: the voice system obeys semantic and syntactic constraints, while IS and its related discourse categories (topics, foci) follows discourse parameters. A morphosyntactic analysis of the distinctive features of topics and focusing devices is conducted, showing how pragmatics maps onto syntax. Among focusing devices are evidential in situ foci, which recruit two evidential forms, han(=tu) and sa(=tu), and cleft narrow foci (informative or restrictive). Evidential foci also express stance and do not trigger any syntactic restructuring, in contrast with extracted cleft foci. As in other Formosan languages, a cleft core argument must be the syntactic pivot of the clause containing the presupposition, while there is no such constraint for cleft peripheral arguments and for adjuncts, which trigger a complement clause. [ABSTRACT FROM AUTHOR]

This paper argues that Neeleman & Szendroi's (2007) radical pro drop analysis in terms of pronominal morphology cannot completely capture the phenomenon of argument drop in six Formosan tongues, Amis, Paiwan, Puyuma, Bunun, and two Atayal dialects. We propose that the topic-oriented analyses of Huang (1984a, 1984b, 1987, 1989) and Sato (2011), which take discourse topic as a major licenser of argument drop, should be adopted to account for the missing arguments in these languages. [ABSTRACT FROM AUTHOR]

This study investigates the development of the multifunctional lexeme aya in Paiwan, a Formosan language spoken in the southern part of Taiwan. Though the lexeme aya has long been identified as having utterance-related usages, few studies were devoted to examining the morphosyntactic behavior of the lexeme, nor did any of these studies provide a systematic account of the relations and the development of its various usages. Based on data collected from fieldwork as well as from One hundred Paiwan texts, this study identifies two major functional domains of aya, represented by SAY verb aya and demonstrative aya: these two instances of aya display distinct morphosyntactic behavior in addition to their diverse functions. SAY verb aya is grammaticalized as a quotative marker, a complementizer-like marker, and an evidential marker, while demonstrative aya develops into a stance marker. Despite these two seemingly unrelated development paths, SAY verb aya and demonstrative aya do share some features in common. It is proposed that Paiwan SAY verb aya, predominantly used in direct quotation of speech events, is inherently encoded with two important features, mimeticity and deicticity, and, therefore, provides a mechanism from which demonstrative aya evolved. This study has theoretical and typological implications. Theoretically, it points out one important yet frequently overlooked fact: that SAY verbs, especially those predominantly used in direct quotation of speech events, are intrinsically mimetic and deictic and, therefore, may function as the origin from which demonstrative elements derive. This study also provides insight for further typological studies related to SAY verbs in Formosan languages and other languages in the world at large. [ABSTRACT FROM AUTHOR]

The aim of this paper is twofold. We first provide a reassessment of the voice systems of Kanakanavu and Saaroa, two Austronesian languages spoken in southern Taiwan), and in particular show that their voice forms that were previously identified as patient focus (patient voice or undergoer voice: patient), locative focus (locative voice or undergoer voice: locative), and beneficiary/instrument focus are actually nominalized forms. Our findings allow us to reconsider the position of Kanakanavu and Saaroa among the Formosan languages. We take as a point of departure Ross's (2009) subgrouping hypothesis whereby Proto-Austronesian (PAN) includes four primary offshoots: Tsou, Rukai, Puyuma, and all other Austronesian languages. This later subgroup, dubbed Nuclear Austronesian, is identified on the basis of the "nominalization-to-verb" innovation, whereby the PAN affixes *-en, *, *-an, *Sa-/*Si-, which were only used in forming nominalizations, were expanded to encode verbal usage in Proto-Nuclear Austronesian. Under this hypothesis, Kanakanavu and Saaroa are both viewed as being Nuclear Austronesian languages. We try to map our findings along with Ross's (2009) reconstruction, and in so doing we are led to place Kanakanavu and Saaroa higher up in Ross's (2009) subgrouping tree and to propose a new hypothesis for the higher phylogeny of the Austronesian languages. [ABSTRACT FROM AUTHOR]

The article discusses the connections of Austronesian. Topics discussed include homeland of Austronesian (AN) located in Taiwan, China, morphological evidence for Sino-Tibetan-Austronesian (STAN), cultural vocabulary comparisons AN monosyllabic root is taken in isolation as a comparandum, and comparison of Old Chinese and Proto-Austronesian.

This paper explores how those with same-sex desire in Taiwan reconfigure Mandarin linguistic practices as western constructions of gay culture are now circulating globally. Ethnographic methods are employed to examine the linguistic stylization that enables Taiwanese gay men use to formulate identities. The nascent emergence of identity markers is first introduced. This study shifts its focus to the context in the study of gay language to avoid circularity. Three primary linguistic stylizations widely practiced among gay cliques are retrieved as (1) kinship soap opera; (2) geisha memoir; (3) celebrity stardom to illuminate the relationship between language and subjectivity. This study illustrates how these vernacular utterances and linguistic stylizations producing collective resources of resistance, accommodation and pleasure are discursively constructed and represented in a localized site in the intersection of an increasingly globalized queer culture. [ABSTRACT FROM AUTHOR]

Linguistic evidence for a knowledge of iron that predates the archaeological evidence for iron technology has had a checkered history in Austronesian linguistics over the past four decades. This squib reevaluates five comparisons first proposed by the writer in 1976, and discards three of them. Based on internal Formosan evidence from languages belonging to different primary branches of the family, it then draws attention to two new comparisons relating to iron that are not likely to be due to diffusion, and raises the question once more whether a knowledge of iron might have preceded iron-working in the Austronesian world by several millennia. [ABSTRACT FROM AUTHOR]

This paper studies preverbal objects in Taiwan Southern Min (TSM), which are defined as NPs that occur after the subject, before the verb, and are subcategorized for by the verb. This paper adopts five syntactic tests taken from Rizzi (1997) to judge the status of preverbal NPs in TSM, topic or focus. First: A topic (but not a focus) may contain a resumptive pronoun in the comment clause. Second: A topic does not result in weak crossover effect, yet the focus does. Third: More than one topic is allowed in one sentence, while the number of the focus is limited to one. Fourth: Bare quantificational elements cannot be topics, yet they can be a focus. Fifth: A wh-operator is compatible with a topic, but not with a focus. Three more factors discussed in the literature on preverbal NPs in Mandarin Chinese are also considered. First: In a topic construction, the VP can be questioned, while in a focus construction, it cannot. Second: A focus can be indefinite but a topic cannot. Third: The bare NP in the pre-adverbial position functions as a topic and it has to be specific, while the NP in the post-adverbial, preverbal position serves as a focus and it can be nonspecific. The test results show that preverbal NPs in TSM can function as either topic or focus when appropriate contrasts are provided. Moreover, through tests, this paper further distinguishes focuses from contrastive topics, which carry intonational stress as focuses do but display features of topics. [ABSTRACT FROM AUTHOR]

Abstract: Situated in a recently established field of variational pragmatics (), this study investigates one of the under-researched non-Indo-European languages, Chinese, with regard to the influence of macro-social and micro-social factors on compliments. More specifically, the present study focuses on the impact of region, a macro-social variable, and compliment topic, a micro-social factor, on Chinese compliments given by Taiwan Chinese and Mainland Chinese higher education students. Sixty Taiwanese and sixty Mainland Chinese, equally gendered in each group, completed a written discourse completion task consisting of eight content-enriched situations () eliciting compliments. With regard to the impact of region, commonalities emerged between those compliments of Chinese students in Taiwan and Mainland China. Both groups preferred to offer Explicit compliments as well as Implicit compliments in the form of Requests, Assumptions, and Want Statements. Overall, Explicit compliments emerged as the most popular strategy. However, statistically significant differences were identified between the two groups in a few Implicit compliment strategies. Regarding the effect of compliment topic, both Taiwan and Mainland Chinese students utilized several compliment strategies in similar ways across appearance/possession and performance/ability situations. It appears that in most cases, it was compliment topic rather than the variety of Chinese which modulated the compliments by both groups. In addition, the paper suggests that compliments in Taiwan and Mainland Chinese may have undergone a change, possibly influenced by western cultures. [Copyright &y& Elsevier]

The present paper investigates the grammatical properties of interrogative verbs in Kavalan and Amis, two Formosan languages. The interrogative words that denote 'do what', 'what happen', 'do how', '(put/take) where', and '(do) how many/much' in the two languages share the same morphosyntactic distribution as verbs. Their transitivity is correlated with the voice markers affixed to them. The affixation of the agent voice marker to an interrogative root forms an intransitive verb, whereas an interrogative root with the patient voice marker is interpreted as a transitive verb. Some interrogative verbs can occur in a verb sequencing construction where they precede a lexical verb. It is argued that the two verbs in this construction do not form a coordinate structure, but exhibit properties of subordination with the interrogative verb as the main verb. Finally, there are semantic restrictions on the use of interrogative verbs. Kavalan tanian and Amis icuwa '(put) where' show verbal properties only when they question the location of a theme argument in a ditransitive event, but not the location where an event takes place. When Kavalan tani and Amis pina/hakuwa '(do) how many/much' are used as verbs, the question always implies that the quantity of the questioned argument will or might change eventually. [ABSTRACT FROM AUTHOR]

This paper is a follow-up investigation of conjunctive reduction in two Formosan languages, Mayrinax Atayal and Southern Paiwan. We examine phenomena surrounding the two-way grammaticalization of linkers—that is, adverbialization and complementation—and show how it works to shape the sentence structure of the two languages. More specifically, it is argued that the Mayrinax linker 'i' licenses modal/evaluative construals on the complementizer layer, and control/middle construals on the lexical layer. In addition, Southern Paiwan evidences a third type of linker construals on the inflectional layer, where frequency/repetitive/locative expressions are introduced, and permutation around the linker a is allowed. All in all, our comparative study of Mayrinax and Squliq Atayal confirms that even in a language without overtly grammaticalized linkers, conjunctive reduction can still be in action, albeit in its final stage of development. [ABSTRACT FROM AUTHOR]

Abstract: This paper presents evidence from the Austronesian languages of Taiwan (Formosan languages) bearing on the ongoing debate between followers of and : are adverbs adjoined to the clause, with semantically determined ordering, or are they located in functional projections in the backbone of the clause, with the ordering determined by universal hierarchical structure? Here it is shown that while the phenomenon of adverbial verbs in Formosan languages cannot be captured under an adjunction analysis, ordering facts still appear to be semantically, rather than structurally, determined. It follows that these two issues (structure and ordering) must be teased apart. It is claimed that the structure, for Formosan at any rate, must be as proposed by Cinque, while the ordering rather seems to behave as would be expected under Ernst''s analysis. Reconciling these two viewpoints implies a new view of syntactic structure which eliminates the clear distinction between adjuncts and functional projections, and which incidentally also points towards a solution of the innateness problem, suggesting that universal syntactic principles do not encode labelled structure as such, but merely cognitive tools for building structure. [Copyright &y& Elsevier]

A number of previous studies on the English pronunciation of native speakers of Mandarin have discovered many sound features distinct from Standard English, yet none of these analyse how frequently each of these features occurred in learners with different English proficiency levels (Chang, 1991; Chen, 1975; Chung, 2006; Gao, 1995; Lee, 1986). This study focuses on the frequency of such pronunciation features amongst intermediate-level learners of English. To this end, in this study ten Taiwan Mandarin speakers read 1,225 words commonly used in English, and the findings indicate that a total of 11 sound features regularly appeared in the pronunciation of the respondents. Among these, five features have not been described in prior research. In particular, three of these features are identified as those which make Taiwan Mandarin-accented English distinct from other varieties of English. This study details the results and concludes by discussing this Expanding-Circle variety of English from the perspective of World Englishes. [ABSTRACT FROM AUTHOR]

This paper investigates metaphorical euphemisms underlying the categories of RELATIONSHIP and DEATH in three Formosan languages: Kavalan, Paiwan, and Seediq, within the framework of Lakoff and Johnson's Cognitive Metaphor Theory. The term "metaphorical euphemism" is proposed to represent both linguistic and cognitive relations of euphemism and metaphor. A metaphorical euphemism refers to a euphemism that adopts metaphorical mapping of both source and target domains to express the notion of a forbidden domain as a result of conscious choices from pragmatic competence. Given the domains of RELATIONSHIP and DEATH being interconnected socially and culturally, nearuniversal and specific metaphorical euphemisms of the two domains in the three languages are analyzed from the approach of descriptive sociolinguistics. A cross-linguistic comparison of the three Formosan languages with English and Mandarin Chinese is provided and discussed from the perspectives of cognition and culture. [ABSTRACT FROM AUTHOR]

The present study investigates the lexicalization of zhenshi and shizaishi in Taiwan Mandarin, showing how they originate as a fusion of an adverb and the copula/focus marker shi and further develop a non-compositional meaning through talk-in-interaction and function as right peripheral stance markers in speech. Many instances of the copula/focus marker shi in Modern Mandarin have become a word-internal element and merged with an adverb/conjunction to form an X-shi construction semantically similar to the original X. Our targets, zhenshi and shizaishi, are instantiations of this morphological trend. However, through sequence truncation and metonymic inference, the two morphological fusion cases further demonstrate lexicalization as they respectively acquire an evaluative meaning. Our study thus illustrates the emergent nature of grammar/lexicon arising from the interaction among language (syntagmatic co-occurrence), culture (communicative politeness), and cognition (pragmatic inference). [ABSTRACT FROM AUTHOR]

This study investigates a newly emergent Taiwan Mandarin slang term taike and its derivative adjective tai. Taike evokes a cultural stereotype of young Taiwanese adults whose lifestyle, linguistic repertoire, and fashion sense are considered distinctively local and unsuccessfully imitative of current trends, but the term has also been appropriated to index local identity and nonconformity. Drawing on data from 344 questionnaires, this study explores the ways the two terms interact with language ideologies and ideologies concerning localism, cosmopolitanism, and authenticity in Taiwan. Statistical results indicate current attitudes toward the terms in Northern Taiwan tend to be negative, with females, the higher-educated, and the younger evaluating the terms more negatively than their counterparts. Further exploration of the data indicates that the meaning contestation is ultimately associated with the global-local ambivalence prevalent in contemporary Taiwan. [ABSTRACT FROM AUTHOR]

This paper provides an analysis of the ambiguous interpretations of the durative complements in the Hakka verb copying construction, schematically represented as "NP1 V-NP2 V C3", in terms of the constructional approach and the cognitive approach. This study has the following proposals. First, the interpretation of the durative complement hinges crucially on the aspectual properties of the V-NP2 unit. Second, contrary to what is generally assumed, a verb copying sentence may be ambiguous between the continuity reading and the repetition reading and can be viewed as a case of constructional polysemy. Third, such ambiguities arise due to the interaction between the episodic/generic duality of the V-NP2 unit and the duration/lapse of time duality of the durative complement and are a natural consequence of the gestalt effect. Finally, the varying degrees of markedness of such a sentence, influenced by world knowledge, depend mainly on the compatibility of the V-NP2 unit and the durative complement. [ABSTRACT FROM AUTHOR]

In this study, we investigate the discourse-functional properties of the extended intransitive clause (EIC) in relation to other clause types in three Formosan languages: Tsou, Kavalan, and Squliq. We offer evidence, based on tracking behavior of NPs, to show that case-marking in EICs is motivated by a core/oblique distinction, and that the core/oblique distinction arises in a systematic way from recurrent patterns in discourse. The referents of the oblique-marked argument nominals in EICs in these languages are shown to be consistently much less likely to be tracked or to be continuous than either the nominative or the ergative argument of a nonactor voice clause. The data derived from investigating the tracking behavior of noun phrases suggest that the validity of a grammatical transitive/intransitive distinction in these languages is quite robust, and that the distinction can best be correctly discerned by examining the functioning of various argument nominals in discourse. [ABSTRACT FROM AUTHOR]

An important issue in the study of the language of emotion language, from a cognitive perspective, is to investigate how language conceptualizes emotions. Since the 1980s, linguists have been exploring how abstract concepts may be based on conceptual metaphors for bodily, physical concepts. However, conceptual metaphors have been shown not to be a universally preferred strategy in conceptualizing emotion events. Nevertheless, the factor influencing the selectionof metaphorical strategies is underexplored and opaque. This study thus sets out to accomplish two goals. First, I postulate that limitation on the linguistic mechanism, namely, nominalization of emotion terms, may constrain language users’ selection of metaphorical strategies in conceptualizing emotion events. My results show that languages having emotion nouns are found to have comparatively rich metaphorical expressions for emotions. To the contrary, in the languages where the linguistic mechanism to encode abstract emotion concepts is inadequate or unavailable, the metaphorical strategy is dispreferred. Second, as emotion events are inherently events of causality, and are characterized by deploying event participants, I examine how grammatical constructions, that is, recurrent syntactic patterns with their associated functions, are made available by the conventional resources of a language, and I investigate how language users in their daily communication conceptualize emotion events and portray participant relations by choosing different grammatical constructions. [ABSTRACT FROM AUTHOR]

Ross (2009) proposes the Nuclear Austronesian hypothesis, according to which the Formosan languages Puyuma, Rukai, and Tsou are each probably a primary branch of Austronesian and all Austronesian languages other than these three belong to a single, Nuclear Austronesian, branch defined by the nominalization-to-verb innovation originally proposed by Starosta, Pawley, and Reid (1981, 1982) for Proto-Austronesian itself. Sagart (2010) argues that there is evidence that Puyuma has also undergone the nominalization-to-verb innovation and is accordingly not a primary branch of Austronesian. In this short paper we show that Sagart's evidence is based on misanalyses of Puyuma data and that these data do not reflect the nominalization-to-verb innovation. Sagart's argument against the Nuclear Austronesian hypothesis does not stand up to closer scrutiny. [ABSTRACT FROM AUTHOR]

This paper presents a critique of Sagart's (2004) classification of the Formosan languages. Sagart proposes a subgrouping based on a set of innovations in the numeral systems of the Formosan languages. These innovations entail that the higher phylogeny of Austronesian is much more hierarchical than in other subgrouping accounts (e.g., Blust 1999, Ross 2009). According to Sagart, the innovated numeral forms can be derived from the complex numerals of the Formosan language Pazeh. A number of arguments are presented that call this subgrouping account into question. There are reasons to assume that the proposed numeral derivations are highly unlikely, and that similarities between the complex numerals of Pazeh and the innovated forms in other Formosan languages are due to chance. Furthermore, language contact might account for the pattern of numeral innovation. Finally, the migration pattern that is entailed by Sagart's numeral innovations does not take into account possible back and forth migrations of the Formosan tribes. [ABSTRACT FROM AUTHOR]

This paper has two main goals. First, I demonstrate how the peculiar morphosyntactic properties of the theme NPs in Isbukun Bunun existential sentences can be derived under an expletive account. The properties under investigation include structural Case marking, the emergence of definiteness restrictions, and the resistance of theme NPs to Ā-extraction. Second, I argue for a small clause analysis of existential sentences in Isbukun Bunun. In particular, based on word orders, island effects, the comparison with possessive constructions, and the coordination facts, I argue that the existential predicate in Isbukun Bunun selects a small clause as its sole internal argument. I also discuss how the proposal developed for Isbukun Bunun might generalize to, and shed light on, the impersonal constructions in natural languages. [ABSTRACT FROM AUTHOR]

This paper presents a structural statistical machine translation (SSMT) model to deal with the data sparseness problem that occurs as a result of the necessarily small corpus to translate Chinese into Taiwanese Sign Language (TSL). A parallel bilingual corpus was developed, and linguistic information from the Sinica Treebank is adopted for Chinese sentence analysis. The Synchronous Context Free Grammar (SCFG) was adopted to convert a Chinese structure to the corresponding TSL structure and then extract a translation memory which comprises the thematic relations between the grammar rules of both structures. In structural translation, the statistical MT (SMT) approach was used to align the thematic roles in the grammar rules and the translation memory provides the reference templates for TSL structure translation. Finally, the agreement information for TSL verbs was labeled for enriching the expressiveness of the translated TSL sequence. Several experiments were conducted to evaluate the translation performance and the communication effectiveness for the deaf. The evaluation results demonstrate that the proposed approach outperforms a baseline statistical MT system using the same small corpus, especially for the translation of long sentences. [ABSTRACT FROM AUTHOR]

Kavalan is an endangered Formosan plains-tribe language still actively spoken in daily life by fewer than one hundred native speakers. This paper investigates reduplication in Kavalan in terms of pattern and meaning. Having reviewed the relevant literature (e.g., Li 1982, 1996; Lin 1996; Chang 2000; Li and Tsuchida 2006; Lee 2007), 1 propose that Kavalan reduplication can be generalized into the following patterns: (1) disyllabic reduplication; (2) monosyllabic reduplication; (3) Ca-reduplication; and (4) lexicalized reduplication. Realization of reduplicants in Kavalan is phonologically driven. A flowchart is used to indicate the process of reduplicant formation. The discussion on the meanings is based on a morphosemantic perspective (cf. Kiyomi 1993), dividing reduplication into verbal and nominal types. The former manifests the meanings of iteration, continuation, intensity, intensification (with sia- or saa-), attenuation, reciprocity (with sim-), and pretense (with sa-); the latter has meanings of plurality, distributiveness, indefiniteness, and smell (with su-). Ca-reduplication is used to indicate intensification (e.g., the increasing quality of color terms), and lexicalized reduplication is often found with onomatopoetic verbal roots and names of fauna, especially birds and insects, and flora. [ABSTRACT FROM AUTHOR]

Most previous studies on serial verb constructions (SVC) in Austronesian languages have focused almost exclusively on two-verb constructions. In this study, we examine serial verb constructions involving three or more verbs in four Formosan languages-Kavalan, Saisiyat, Squliq Atayal, and Tsou-focusing our attention on the type of verb serialization that observes the same-subject condition. Action or motion verbs are found to be the only attested final verb in an SVC in these languages. General strategies for triple verb serialization result when any two types of the nonfinal verbs in a two-verb SVC-namely, modal, emotion, evaluative/manner, motion, or "adverbial" verbs-occur before the final action/motion constituent verb in some specific linear order, while Tsou and Kavalan recruit, in addition, one or two types of "sentential adverbial" verbs and degree or quantifier verbs in a language-specific triple verb serialization strategy. A consistent ordering pattern of the various types of verbs is found in all of the languages studied. Quadruple verb serialization, found only in Tsou and Kavalan, results when certain subtypes of "adverbial" verbs occur in combination with certain other subtypes. Our findings are also consonant with a central tenet in theories of construction grammar, that constructions in languages are not broad syntactic templates, but are typically lexically skewed schemas and formulas. [ABSTRACT FROM AUTHOR]

This paper investigates the diachrony of the verbs of volition ai and beh in Southern Min (SM). The data include historical SM texts (16th-19th century), folk songs in the late 19th to early 20th century, and contemporary natural Taiwanese Southern Min conversations. The results mainly mark two stages of the development of ai and beh: historical (before 1900) and contemporary SM. Since the 16th century (Ming Dynasty), ai has been used as a verb of volition indicating love, intention or hope. As for historical ai, it marked future, specialized however for predicting an adverse future. In contemporary SM, a sense of necessity emerged with the use of ai denoting general agreement among people. Seven senses are attributed to historical beh: want/intention/hope, future, excessiveness, necessity, proximity and conditionality. In contemporary SM, the indication of the want of an entity has become a less preferred use, and necessity is only preserved when indicating puzzlement and helplessness. The diachronic developments of ai and beh demonstrate an interplay of metaphorization and metonymization (Traugott and Dasher 2002:27). Metaphorization contributes first to the semantic shift of ai and beh from "to want; to love" to "to intend to" and a concurrent categorical change from lexical verb into auxiliary, and second, to the evolution of future. Metonymization activates an even wider range of uses and meaning change, including the pragmatic strengthening of interpretations such as excessiveness (beh), necessity (ai, beh), proximity (beh) and conditional (beh), and prediction of an undesirable future (ai). [ABSTRACT FROM AUTHOR]

Postlexical prosodic phenomena in Austronesian languages have received relatively little attention, and consequently their patterns remain unknown. This paper aims to bridge this gap by investigating how syntactic structure interacts with postlexical prosodic phenomena in Saisiyat, an endangered language spoken in Taiwan. Several significant findings are made. First, Saisiyat sentential fundamental frequency (F0) patterns are based largely on its original lexical-level word F0 contour. The accent of a content word in sentences usually falls on its ultimate syllable, while trisyllabic or quadrisyllabic content words may sometimes undergo postlexical accentual modifications such as accent spreading, accent fronting, and accent adding. Function words, in contrast, play a role of interpolation as an intermediate site in bridging the F0 of their preceding and following syllables. Second, a yes-no question exhibits substantial prosodic modification by influencing its word preceding sentence-final interrogative particle aj, as compared with its counterpart in a declarative sentence. Third, agents in agent-focus sentences and patients in patient-focus sentences demonstrate higher values with respect to F0 peak, mean F0, and mean intensity. We provide typological explanations for these findings and explore the theoretical implications of postlexical prosodic patterns of Formosan languages in Taiwan. [ABSTRACT FROM AUTHOR]

Information technologies have now matured to the point of enabling researchers to create a repository of language resources, especially for those languages facing the crisis of endangerment. The development of an online platform of corpora. made possible by recent advances in data storage, character-encoding and web technology, has profound consequences tot the accessibility, quantity, quality and interoperability of linguistic field data. This is of particular significance for Formosan languages in Taiwan, many of which are on the verge of extinction. As a response to the recognition of this burgeoning problem, the key objectives of the establishment of the NTU Corpus of Formosan Languages aim to document and thus preserve valuable linguistic data. as well as relevant ethnological and cultural information. This paper will introduce some of the theoretical bases behind this initiative, as well as the procedures, transcription conventions. [ABSTRACT FROM AUTHOR]

This paper discusses the issue of how coordinate structures evolve into subordinate structures in both syntactic and semantic terms. I call this type of process "conjunctive reduction." It is well established in the literature on Chinese historical syntax that some modifier-head and verb-complement compounds actually derive from coordinate structures in Ancient Chinese. Based on this finding, I suggest that a similar process is also at work in Formosan languages, but on a quite different scale. That is, while Chinese encodes conjunctive reduction in compounding morphology, the same process involves full-fledged syntactic operations in Formosan languages. I propose that there are two general directions of conjunctive reduction. On the one hand, the coordinator may become a modifier marker, where the first conjunct becomes a marked adverbial; then the modifier marker may disappear completely, making the first conjunct an unmarked adverbial. I call this "adverbialization." On the other hand, the coordinator may become a complementizer, introducing either an infinitive complement or an adverbial adjunct such as a conditional or temporal clause. I take Squliq Atayal, Tsou, and Amis to represent the Northern, Tsouic, and Paiwanic groups respectively, which in turn points to the existence of a protolanguage with extensive coordinate construals along the line of Neo-Davidsonian semantics, very much like Ancient Chinese. [ABSTRACT FROM AUTHOR]

Very few verbs in Formosan languages are known to carry a full set of voice affixes, but it remains unclear how and to what extent the lexical meaning of a verb influences constructional variability. In this study we sort out where the gaps may lie, and probe the complex and interwoven nature of the relationships between the semantic roles of nominative arguments and various non-Actor Voice clauses in Tsou. We argue that voice forms are not automatic mechanical processes based on lexical selection alone, and we offer new insights that show that both lexical and constructional meaning interact in licensing Benefactive Voice (BV) clauses. The argument is also made that two types of BV clauses must be distinguished, causativized and noncausativized. Verbs in noncausativized BV clauses "subcategorize" for nominative NPs that would normally function as "noncore" or peripheral arguments in a European language like English, which thus constitutes important evidence that Tsou is a language that does not grammatically make the core/oblique distinction. Causativized BV clauses, on the other hand, have the effect of bringing in an extra argument to the source clauses and marking it with oblique case, while retaining the case markings on the nominal arguments in the source clauses, violating the well-known causee accessibility hierarchy. [ABSTRACT FROM AUTHOR]

This work proposes a novel approach to translate Chinese to Taiwanese sign language and to synthesize sign videos. An aligned bilingual corpus of Chinese and Taiwanese Sign Language (TSL) with linguistic and signing information is also presented for sign language translation. A two-pass alignment in syntax level and phrase level is developed to obtain the optimal alignment between Chinese sentences and Taiwanese sign sequences. For sign video synthesis, a scoring function is presented to develop motion transition-balanced sign videos with rich combinations of intersign transitions. Finally, the maximum a posteriori (MAP) algorithm is employed for sign video synthesis based on joint optimization of two-pass word alignment and intersign epenthesis generation. Several experiments are conducted in an educational environment to evaluate the performance on the comprehension of sign expression. The proposed approach outperforms the IBM Model2 in sign language translation. Moreover, deaf students perceived sign videos generated by the proposed method to be satisfactory. [ABSTRACT FROM AUTHOR]

In spite of a widening acceptance of attenuated retroflexed initials (zh-, ch-, sh-) in Taiwan Mandarin today, there is a parallel movement in seemingly the opposite direction: a growing use of the retroflexed initials in certain contexts. A conflict between the two trends often surfaces in the form of hypercorrection, that is, incorrect substitution of the retroflexed initials for the corresponding dental initials (z-, c-, s-). Labov (1973) observed a trend toward a similar kind of phonetic hypercorrection in New York City English, mainly among the upwardly-aspiring lower middle class. Though this group is also especially susceptible to the use of hypercorrect forms in Taiwan, people in all walks of life with all levels of education have been observed to use hypercorrect forms. This demonstrates, first, that the textbook forms of the retroflexed vs. dental initials are learned imperfectly by a wide spectrum of speakers of Taiwan Mandarin; second, that the retroflexed initials retain a certain cachet in marking speech as more prestigious and authoritative; and third, that retroflexion, hypercorrect or otherwise, has for many people taken on the function of simply marking formal discourse, in addition to its use for disambiguation, highlighting, and stylistic effect. [ABSTRACT FROM AUTHOR]

This is a general survey of cardinal numerals in Formosan languages. Most languages distinguish between human and nonhuman numerals, not only in cardinal numerals, but also for terms that have to do with number, such as ‘how many/much’, ‘many/much’, and even ‘few/little’. Some languages have a third set of numerals as used in counting, different from both human and nonhuman numerals. Tables of the numerals 1–10 in the still extant Formosan languages are given in the appendixes. Most Formosan languages retain a decimal system, although a few numerals may have been modified in some of the languages. Pazih is the only language that has nearly a quinary system. Numerals may be derived from other numeral stems by addition, subtraction, or multiplication, and some are unique to Formosan languages. Numerals may function either as nouns or verbs, depending on their syntactic position, and they may appear in simple or derived form. [ABSTRACT FROM AUTHOR]

Causatives have been subjected to intensive scrutiny by linguists in recent years. Cross-linguistic studies suggest that the formation of causatives reflects the real world perception of cause-result relations, though some studies contradict this finding. The aim of this study is to explore the underlying principles that determine the way an event is encoded linguistically in Saisiyat, a Formosan language. Following Croft's model of idealized single events, we promote the study of causatives to the discourse level. The results show that the iconicity of language is reflected in the ordering of linguistic elements as well as their grammatical integrity. [ABSTRACT FROM AUTHOR]

This article analyses the functions and meanings of the reflexes of the Proto-Oceanic (POc) prefix *paRi- in New Caledonian languages and, based on comparisons with other Austronesian languages, seeks to account for their broad polyfunctionality and polysemy in New Caledonia. POc *paRi- originates from Proto-Austronesian and Proto-Malayo-Polynesian *paR- , as evidenced by their reflexes in various Formosan and Malayo-Polynesian languages. In Formosan languages, these prefixes are reciprocal markers and their combination with root reduplication marks plurality of reciprocal participants. In Malayo-Polynesian languages, the prefixes have a greater variety of functions and meanings: they are basically intransitive actor focus morphemes which also express middle voice and reciprocity as well as plural relationship, intensive, and frequentative meanings. In Oceanic languages, the reflexes of POc *paRi- express symmetrical, plural relationship and lower transitivity. POc *paRi- could also combine with two suffixes (POc *-i , *-aki ) or with reduplication to express some other meanings (intensive, iterative, distributive, dispersive), as attested in various Oceanic languages. Thus, originally reciprocal prefixes in Formosan languages evolved into more polysemous morphemes in Malayo-Polynesian languages and developed in more or less conservative or innovative ways in Oceanic languages. In the northern New Caledonian languages, the decay of some of their combined suffixes and of reduplication probably triggered the conflation of these various functions and meanings into the preserved and productive prefixes. [ABSTRACT FROM AUTHOR]

We introduce the newly developed Formosan Language Archive and show how it has been built up through examples drawn from Rukai, a Formosan language spoken across southern Taiwan and including six main dialects (Mantauran, Maga, Tona, Budai, Labuan, and Tanan) that exhibit great variation. After displaying the layout of the archive, we explain how texts and sound files are recorded and digitalized, how words are tagged, and what the purpose of the search system is. Last, we compare the Formosan language archive to other well-established language archives and show how and why we have adopted a layout that is somewhat different but enables us to capture, through thorough linguistic analysis, the variations displayed in the Rukai dialects (and in Formosan languages in general). [ABSTRACT FROM AUTHOR]

Proto-Oceanic had two distinct sets of clitic pronouns, with members of one set expressing the S (‘subject’ of intransitive sentences), while the members of the other set expressed the A (‘agent’ of transitive sentences). Clitic pronouns expressing O (‘object’) observed in some Oceanic languages today are later innovations, rather than reflexes of Proto-Oceanic ‘object’ clitic forms. The main evidence comes from a comparison of what is reconstructible for Proto-Extra Formosan, a parent language of Proto-Oceanic, and Proto-Polynesian, one of the daughter protolanguages of Proto-Oceanic. [ABSTRACT FROM AUTHOR]