Re-evaluations of the category of deponency in recent years have been the leading cause of a paradigm shift taking place in studies on the ancient Greek voice system, opening up new avenues for further remodelling. The present study contends that verbal voice operates according to an ergative two-voice system, active and middle-passive, producing two contrastive roles the subject plays in a clause. Within a nominativeaccusative alignment patterning, which marks out transitive operations of a clause, ergative functions centered on verbal voice are present in the language’s verbal morphology and syntax. An ergative view of voice specifies different transitive participant roles and focuses on the affected element of the clause that realizes or actualizes the verbal process. Clearer expression of the subject’s function in the clause occurs by distinguishing between two opposing roles: the subject functioning either as realization of the process or as cause of the process. Two basic and contrastive roles of the subject, therefore, mean that two semantic domains for voice are operational in the language system network despite the occurrence of three morphological forms in the aorist and future tense-forms. The middle and passive uses, together comprise the middlepassive voice, and the active voice comprises the other voice domain. Middle and passive functions share the common feature of subject-affectedness, but middle uses occur when there is a feature of internal agency in addition to the subject actualizing the verbal process. Passivity occurs when the subject actualizes the verbal process with an added feature of external agency to the clause. Moreover, passivity takes place through specific grammatical constructions within the middle-passive voice that operate as agentive augmentations (specified or not) of a middle-passive clause type. This is frequently expressed using the so-called, ‘passive marker,’ -(θ)η, that was encroaching upon middle forms during this stage of the language and gradually expanding its range of function in the New Testament writings. / Thesis / Doctor of Philosophy (PhD)

This dissertation provides a formal and functional analysis of grammatical voice in Sasak, an Austronesian language spoken in Eastern Indonesia. The research addresses two primary questions, which are (1) how does Sasak clause structure and morphosyntax vary across dialects? and (2) what shapes speakers’ syntactic production, namely grammatical voice choices? Answers to these questions are pursued via elicitation data, a corpus analysis, and results of two language production experiments. The first part of the dissertation examines how Sasak dialects differ syntactically and morphosyntactically. Data from embedded clauses, clitics, and possessive pronominal clitics are used to argue that that Central Sasak maintains two distinct transitive clause types despite the lack of the overt morphological contrast found with transitive verbs in Eastern Sasak. These data also support prior arguments (Davies, 1993; Guilfoyle, Hung, & Travis, 1992; Shibatani, 2008) that Indonesian languages have either two grammatical subject positions, or both a subject and grammatical topic position in the case of Sasak. Many Austronesian languages spoken on Indonesia’s Java Island and surrounding islands share a cognate nasal prefix that is generally found in the presence of preverbal actors (Arka, 2009; Davies, 2005; Sneddon, 1996). This dissertation presents data from three Sasak dialects that show how multiple, morphologically distinct nasal prefixes in Sasak dialects (also noted by Austin, 2012) correlate with two syntactic facts: first, what argument may be extracted out of vP; and secondly, whether or not the lexical verb projects an internal argument. These facts are accounted for in a Minimalist framework (Chomsky, 1993, 2001) by permitting variation to target single features on syntactic heads (as proposed by Aldridge, 2008). The second half of the dissertation investigates what factors shape speakers’ grammatical voice choices. Speakers’ production patterns can clearly be understood as shaped by the structural properties of their specific language(s), and this is also true in Sasak. However, what about when multiple word orders and voice choices are possible? When languages allow for syntactic options, are there universal non-syntactic constraints that exert influence on the production and syntactic coding choices? This dissertation explores potential universal biases identified in literature that has grown out of Bock and Warren’s (1985:50) work on Conceptual Accessibility, or the “ease with which the mental representation of some potential referent can be activated in, or retrieved from, memory”. The specific biases examined for Sasak in the current work are Discourse Topicality (Givón, 1983), animacy (Branigan, Pickering, & Tanaka, 2008), and noun phrase length (MacDonald, 2013; Tanaka, Branigan, McLean, & Pickering, 2011). Results of a corpus analysis are combined with data from two production experiments, and show that both animacy and topicality affect voice selection in Sasak. Specifically, [+animate] and [+topical] noun phrases are produced earlier in a sentence, thereby affecting the grammatical voice produced. Also, Sasak speakers exhibit a ‘long before short’ bias (i.e., placing longer noun phrases before relatively shorter ones in utterances), affecting voice selection as well. Contextualized in cross-linguistic data, this supports the argument made in this dissertation that the cognitive effect of the semantic richness and salience of longer nouns is relative to the speaker’s stage in planning and producing an utterance.

本研究旨在探討雅美語雙及物結構是否也有與英文，及其他語言如日語、希臘語、國語，一樣有與格轉換的語言現象。雅美語有一特別的強調系統可將名詞組移到主詞的位置上。在探討雅美語的雙及物結構前，必須先探討幾個問題。首先，研究名詞組的格位標記是否會因動詞的性質（動詞的及物性、動詞的論元結構）不同而有所改變，其次，提出論證證明強調系統並不適用於形容雅美語特殊的移位系統，而語態系統較能更進一步的形容此一名詞移位現象。最後，提出雅美語並不是一個作格語言，而是一個valency-neutral的語態系統。解決這些問題後，發現雅美語中也有類似英文中的與格轉換的語言現象，此一發現也驗證了Harley的提案，擁有“有”動詞的語言就會有與格轉換的現象。 / The aim of the present study is to examine the existence of dative alternation in Yami, a language with a rich case marking system, that is similar to Japanese and Greek which are both reported to have dative alternation, as well as very unique ‘focus’ systems that can promote any argument into the subject position. Several issues have to be addressed: first, the case marking on the nominal is investigated in four most commonly observed ‘focus’ constructions from various aspects including degree of transitivity, thematic structure, and event classes,…etc; second, the term ‘focus’ is misleading and is identified as ‘voice’; third, arguments against Yami as an ergative language and supports for a valency-neutral voice system are provided.　Once these basic linguistic properties have been clarified, an examination of trivalent verbs shows positive evidence of dative alternation in Yami. Dative alternation is found with the two trivalent Yami verbs meaning ‘distribute out/give out’ and ‘mail’. This finding is in accordance with Harley’s proposal of the co-existence verbal HAVE and dative alternation, and also suggests that dative alternation is not a language-specific property.