Ngalakgan is an Australian language of the Gunwinyguan family, spoken fluently by just a few people in the mid Roper River area of the Top End. The thesis is a description and examination of the phonology, prosody, and morphology of Ngalakgan, based on several years of fieldwork. Ngalakgan is a language with a rich inventory of classically Gunwinyguan morphological features, including noun class agreement for all major and some minor word classes, compounding of both nouns and verbs, and a rich array of modifying and inflectional prefixes and suffixes. In Ngalakgan, there is a distinction between two kinds or 'levels' of morphology: 'root'-level and 'word'-level. Root-level morphology is lexicalised and unproductive. It is restricted to the tense/aspect/mood inflection of the small closed class of 'finite' verb roots, and to the large closed class of compounds of these roots. Word-level morphology is productive, and includes almost all prefixes, all (non-tensed) suffixes and all clitics. Only word-level structure is consistently reflected in prosodic structure; forms which are complex only at the root-level are treated as prosodic units. I show that all word-level morphemes constitute prosodic domains: every word-level stem, affix and clitic potentially begins a new domain for metrical foot structure. Geminates and glottal stops are over-represented at morpheme boundaries in complex words. In addition, they are subject to complex, non-local alternations with simple stops and zero, respectively, in Ngalakgan and related languages. The alternations are conditioned by preceding geminates and voiceless obstruent clusters, as well as by prosodic and morphological structure. I propose that voiceless obstruent clusters constitute 'boundary signals' to morphological structure, in a similar fashion to stress and, like stress, are 'licensed' by the organisation of intonation. Ngalakgan displays a quantitive-sensitive stress system in roots which is apparently unique to languages of this area. Heavy syllables in Ngalakgan are those which are articulatorily and perceptually complex: those in which the coda is followed by a consonant with a distinct place of articulation. Geminates, homorganic nasal+stop clusters and glottal stops interact with this distinction in ways which are not predicted by current prosodic theories.

Results of a production experiment on the placement of sentence accent in German are reported. The hypothesis that German fulfills some of the most widely accepted rules of accent assignment— predicting focus domain integration—was only partly confirmed. Adjacency between argument and verb induces a single accent on the argument, as recognized in the literature, but interruption of this sequence by a modifier often induces remodeling of the accent pattern with a single accent on the modifier. The verb is rarely stressed. All models based on linear alignment or adjacency between elements belonging to a single accent domain fail to account for this result. A cyclic analysis of prosodic domain formation is proposed in an optimality-theoretic framework that can explain the accent pattern. Japanese <i>wh</i>-questions always exhibit focus intonation (FI). Furthermore, the domain of FI exhibits a correspondence to the <i>wh</i>-scope. I propose that this phonology-semantics correspondence is a result of the cyclic computation of FI, which is explained under the notion of <i>Multiple Spell-Out</i> in the recent Minimalist framework. The proposed analysis makes two predictions: (1) embedding of an FI into another is possible; (2) (overt) movement of a <i>wh</i>-phrase to a phase edge position causes a mismatch between FI and <i>wh</i>-scope. Both predictions are tested experimentally, and shown to be borne out.

In this paper we review the current state of research on the issue of discourse structure (DS)/information structure (IS) interface. This field has received a lot of attention from discourse semanticists and pragmatists, and has made substantial progress in recent years. In this paper we summarize the relevant studies. In addition, we look at the issue of DS/ISinteraction at a different level - that of phonetics. It is known that both information structure and discourse structure can be realized prosodically, but the issue of phonetic interaction between the prosodic devices they employ has hardly ever been discussed in this context. We think that a proper consideration of this aspect of DS/IS-interaction would enrich our understanding of the phenomenon, and hence we formulate some related research-programmatic positions.

In this dissertation, I examine the marking of focus and givenness in Nte?kepmxcin (Thompson River Salish). The focus is, roughly, the answer to a wh-question, and is highlighted by the primary sentential accent in stress languages like English. This has been formalized as the Stress-Focus Correspondence Principle. Given material is old information, and is de-accented in languages like English. Nte?kepmxcin is a stress language, but marks focus structurally. However, I argue that the structure has a prosodie motivation: the clause is restructured such that the focus is leftmost in the intonational phrase. It follows that Salish focus structures lack the special semantics that motivates the use of English structural focus (clefts). As a theoretical contribution, I show that the Stress-Focus Correspondence Principle does not account for focus marking in all stress languages, nor does the "distress-given" generalization account for the marking of given information. This is because focus surfaces leftmost, while the nuclear stress position is rightmost. Instead of "stress-focus", I propose that alignment with prosodie phrase edges is the universally common thread in focus marking. This mechanism enables listeners to rapidly recover the location of the focus, by identifying coarse-grained phonological categories (p-phrases and i-phrases). In Thompson River Salish, the focus is associated with the leftmost p-phrase in the matrix intonational phrase. The analysis unifies the marking of focus across languages by claiming that focus is always marked prosodically, by alignment to a prosodie category. The study combines syntactic analysis of focus utterances with their phonetic realization and semantic characteristics. As such, this dissertation is a story about the interfaces. This research is based on a corpus of conversational data as well as single sentence elicitations, all of which are original data collected during fieldwork. The second contribution of this dissertation is thus methodological: I have developed various fieldwork techniques for collecting both spontaneous and scripted conversational discourses. The empirical contribution that results is a collection of conversational discourses, to add to the single speaker traditional texts already recorded for Nte?kepmxcin.

Are subjects produced differently based on the transitivity of the clause they are embedded in? Based on data from a narrative reading experiment, it is shown that transitive subjects are produced with a lower f0 than intransitive subjects and that this difference is statistically significant (p∠0.05). It is suggested that the purpose for such a difference originates from a propensity for English speakers to accent new referents, which are common in the object position. By lowering the f0 of the subject, speakers increase the efficacy of an accent on a new object later in the clause. Finally, the read narrative procedure is evaluated for its strict control of stimuli, while also reproducing known intonational phenomena.

This study investigates the perception of vowel lengthening in the tonada cordobesa, a feature of the Spanish spoken in Córdoba, Argentina. Unlike other dialects of Argentine Spanish, lengthening occurs in the pre-tonic syllable (Fontanella de Weinberg 1971; Yorio 1973; Lang 2010) and is believed to be accompanied by a pitch peak (Fontanella de Weinberg 1971). The goals of this dissertation are to determine if duration alone (i.e., without intonational changes) is significant in identifying a speaker’s Cordoba provenance, and to discover what listener features affect perception. A matched-guise methodology presents speech tokens with natural and manipulated pre-tonic vowel durations to Argentine listeners in a dialect identification task. Results show a main effect of speaker region and token type (natural versus manipulated). Shorter durations made Córdoba speakers difficult to identify, reducing accuracy from 59% for natural tokens to 28% for manipulated tokens with shortened pre-tonic syllables. Buenos Aires speakers received the highest identification accuracy for natural tokens (80%) and Tucumán speakers the lowest (43%). Longer pre-tonic vowel durations are associated with a Córdoba identity, regardless of speaker origin and other linguistic cues. Control tokens produced by speakers from Buenos Aires and Tucumán confirmed this effect: these tokens, when manipulated to have a longer pre-tonic vowel, induced the perception of a Córdoba identity. Listener experience is also shown to improve accuracy of dialect identification: listeners of more geographically distant provinces, relative to the speaker’s province of origin, present significantly reduced identification rates. Acoustical analyses of the Cordoba samples confirm pre-tonic lengthening as well as an early peak rise within the stressed syllable, and valley alignment before the onset of this syllable. Pre-tonic, tonic and post-tonic syllable durations are lengthened, resulting in a segmentally unbalanced intonational phrase for which prominence is disproportionately concentrated in these final segments. The durational, intonational, and rhythmic properties make the Cordoba dialect unique among regional lects within Argentina and across the Spanish-speaking world. This research contributes experimental evidence for the prosodic features marking this dialect and supports its saliency and social significance within Argentina. / text

SENĆOŦEN has a set of second-position clitics (2PCs) (‘little’, unstressed elements, such as the first person subject SEN), following the initial prosodic word (full word) of the clause. This thesis, which studies the distribution of the 2PCs, is divided into two parts: a linguistic analysis and a co-authored teaching appendix. In the linguistic analysis, I propose that 2PCs occur following the initial prosodic word as a result of constraints governing the mapping between syntactic and prosodic structure. In the syntax, I propose that SENĆOŦEN 2PCs occupy positions above the prosodic word that ultimately precedes them. However, a preference for ‘strong’ left edges of prosodic constituents (intonational units) results in the violation of the constraint governing linearization of the syntactic structure, allowing the clitics to follow the initial prosodic word. The teaching appendix, developed collaboratively with STOLȻEȽ Elliott, employs concepts from the linguistic analysis in a way that is useful for language learners and teachers. / Graduate / 0290 / mhuijs@telus.net

- / -

Among the many ways in which sounds alternate in the world's languages, changes in the order of sounds (metathesis) are relatively rare. Mutsun, a Southern Costanoan language of California which was documented extensively before the death of its last speaker in 1930, displays three patterns of synchronic consonant-vowel (CV) metathesis. Two of these patterns appear to have remained productive while the language was actively spoken. In stem-deriving metathesis, many disyllabic noun stems ending in a VC string (as well as a few trisyllabic noun stems) alternate with semantically related verb stems ending in a CV string: e.g.,cayic ‘strength’ ~ cayci ‘to be strong’. In reflexive metathesis, a subset of verb stems, which are normally vowel-final in all environments, surface in consonant-final form in the presence of the reflexive suffix –pu and/or the reciprocal suffix -mu, as in kitro ‘to dress, to clothe’ ~ kitorpu ‘to get dressed, to dress oneself’. Finally, in suffix metathesis, the plural and locative suffixes (as well as the desiderative/irrealis enclitic) alternate between CCV and CVC forms depending on whether the preceding stem ends in a consonant or a vowel. Based on data from a large corpus of archival records of the language compiled over a span of more than a century, all three patterns of metathesis in Mutsun appear to defy the types of phonological analysis that have been proposed in the literature to account for metathesis in a variety of other languages. The phonetic and phonological factors claimed to motivate metathesis in other languages, such as misinterpretation of acoustic cues, stress attraction, sonority hierarchies, and positional restrictions, are absent in Mutsun. In this dissertation, I argue that prosodic analyses based on syllable weight and prosodic templates are required to account for Mutsun metathesis. Mutsun stem metathesis in particular has less in common, morphophonologically speaking, with metathesis in other languages than it does with reduplication or templatic morphology.

In this dissertation, I examine the marking of focus and givenness in Nte?kepmxcin (Thompson River Salish). The focus is, roughly, the answer to a wh-question, and is highlighted by the primary sentential accent in stress languages like English. This has been formalized as the Stress-Focus Correspondence Principle. Given material is old information, and is de-accented in languages like English. Nte?kepmxcin is a stress language, but marks focus structurally. However, I argue that the structure has a prosodie motivation: the clause is restructured such that the focus is leftmost in the intonational phrase. It follows that Salish focus structures lack the special semantics that motivates the use of English structural focus (clefts). As a theoretical contribution, I show that the Stress-Focus Correspondence Principle does not account for focus marking in all stress languages, nor does the "distress-given" generalization account for the marking of given information. This is because focus surfaces leftmost, while the nuclear stress position is rightmost. Instead of "stress-focus", I propose that alignment with prosodie phrase edges is the universally common thread in focus marking. This mechanism enables listeners to rapidly recover the location of the focus, by identifying coarse-grained phonological categories (p-phrases and i-phrases). In Thompson River Salish, the focus is associated with the leftmost p-phrase in the matrix intonational phrase. The analysis unifies the marking of focus across languages by claiming that focus is always marked prosodically, by alignment to a prosodie category. The study combines syntactic analysis of focus utterances with their phonetic realization and semantic characteristics. As such, this dissertation is a story about the interfaces. This research is based on a corpus of conversational data as well as single sentence elicitations, all of which are original data collected during fieldwork. The second contribution of this dissertation is thus methodological: I have developed various fieldwork techniques for collecting both spontaneous and scripted conversational discourses. The empirical contribution that results is a collection of conversational discourses, to add to the single speaker traditional texts already recorded for Nte?kepmxcin.