Doctor of Philosophy (PhD) / This dissertation documents the emergence of a new language, Light Warlpiri, in the multilingual community of Lajamanu in northern Australia. It then examines the acquisition of Light Warlpiri language, and of the heritage language, Lajamanu Warlpiri, by children. Light Warlpiri has arisen from contact between Lajamanu Warlpiri (a Pama-Nyungan language), Kriol (an English-based creole), and varieties of English. It is a Mixed Language, meaning that none of its source languages can be considered to be the sole parent language. Most verbs and the verbal morphology are from Aboriginal English or Kriol, while most nouns and the nominal morphology are from Warlpiri. The language input to children is complex. Adults older than about thirty speak Lajamanu Warlpiri and code-switch into Aboriginal English or Kriol. Younger adults, the parents of the current cohort of children, speak Light Warlpiri and code-switch into Lajamanu Warlpiri and into Aboriginal English or Kriol. Lajamanu Warlpiri and Light Warlpiri, the two main input languages to children, both indicate A arguments with ergative case-marking (and they share one allomorph of the marker), but Lajamanu Warlpiri includes the marker much more consistently than Light Warlpiri. Word order is variable in both languages. Children learn both languages from birth, but they target Light Warlpiri as the language of their everyday interactions, and they speak it almost exclusively until four to six years of age. Adults and children show similar patterns of ergative marking and word order in Light Warlpiri. But differences between age groups are found in ergative marking in Lajamanu Warlpiri - for the oldest group of adults, ergative marking is obligatory, but for younger adults and children, it is not. Determining when children differentiate between two input languages has been a major goal in the study of bilingual acquisition. The two languages in this study share lexical and grammatical properties, making distinctions between them quite subtle. Both adults and children distribute ergative marking differently in the two languages, but show similar word order patterns in both. However the children show a stronger correlation between ergative marking and word order patterns than do the adults, suggesting that they are spearheading processes of language change. In their comprehension of sentences in both Lajamanu Warlpiri and Light Warlpiri, adults use a case-marking strategy to identify the A argument (i.e. N+erg = A argument, N-erg = O argument). The children are not adult-like in using this strategy at age 5, when they also used a word order strategy, but they gradually move towards being adult-like with increased age.

Doctor of Philosophy (PhD) / This dissertation documents the emergence of a new language, Light Warlpiri, in the multilingual community of Lajamanu in northern Australia. It then examines the acquisition of Light Warlpiri language, and of the heritage language, Lajamanu Warlpiri, by children. Light Warlpiri has arisen from contact between Lajamanu Warlpiri (a Pama-Nyungan language), Kriol (an English-based creole), and varieties of English. It is a Mixed Language, meaning that none of its source languages can be considered to be the sole parent language. Most verbs and the verbal morphology are from Aboriginal English or Kriol, while most nouns and the nominal morphology are from Warlpiri. The language input to children is complex. Adults older than about thirty speak Lajamanu Warlpiri and code-switch into Aboriginal English or Kriol. Younger adults, the parents of the current cohort of children, speak Light Warlpiri and code-switch into Lajamanu Warlpiri and into Aboriginal English or Kriol. Lajamanu Warlpiri and Light Warlpiri, the two main input languages to children, both indicate A arguments with ergative case-marking (and they share one allomorph of the marker), but Lajamanu Warlpiri includes the marker much more consistently than Light Warlpiri. Word order is variable in both languages. Children learn both languages from birth, but they target Light Warlpiri as the language of their everyday interactions, and they speak it almost exclusively until four to six years of age. Adults and children show similar patterns of ergative marking and word order in Light Warlpiri. But differences between age groups are found in ergative marking in Lajamanu Warlpiri - for the oldest group of adults, ergative marking is obligatory, but for younger adults and children, it is not. Determining when children differentiate between two input languages has been a major goal in the study of bilingual acquisition. The two languages in this study share lexical and grammatical properties, making distinctions between them quite subtle. Both adults and children distribute ergative marking differently in the two languages, but show similar word order patterns in both. However the children show a stronger correlation between ergative marking and word order patterns than do the adults, suggesting that they are spearheading processes of language change. In their comprehension of sentences in both Lajamanu Warlpiri and Light Warlpiri, adults use a case-marking strategy to identify the A argument (i.e. N+erg = A argument, N-erg = O argument). The children are not adult-like in using this strategy at age 5, when they also used a word order strategy, but they gradually move towards being adult-like with increased age.

This dissertation contributes to an ongoing debate on the types of linguistic features which can be acquired in a second language by looking at the multiple learning challenges related to the ergative case system (the appearance of –ne with the subject) in Hindi-Urdu by classroom learners. Some hypotheses in second language research hold that interpretable features (features which contribute semantic information) can be acquired in a second language, whereas uninterpretable features (features which express grammatical information) cannot be easily acquired, if ever. Additionally, hypotheses in second language processing hold that the second language learners are able to process semantic information but not grammatical information. This dissertation investigates at the acquisition process of second language learners of Hindi-Urdu acquiring the uninterpretable ergative case. In Hindi-Urdu, the subject of a sentence appears with the ergative case marker –ne, when the verb is transitive and in the perfective aspect. In my dissertation, I test the validity of the aforementioned hypotheses and investigate the acquisition and acquisitional process of ergative case in L2 Hindi-Urdu by L1 English speakers by analyzing data collected by using an acceptability/grammaticality judgement task, a self-paced reading task and a production task from Hindi-Urdu learners and native speakers.

Le terme d’antipassif, qui s’inscrit depuis quarante ans dans le système des langues ergatives, désigne une construction intransitive ayant pour caractéristique générale la destitution du patient. Cette thèse remet en question l’opinion traditionnelle selon laquelle les constructions antipassives sont identifiées exclusivement dans les langues à alignement ergatif, et non dans les langues à alignement accusatif. Étant donné qu’une certaine proportion de langues ergatives utilise pour dériver l’antipassif le morphème polysémique réfléchie et/ou réciproque, dans cette étude nous nous sommes intéressée aux langues accusatives dont la marque antipassive présente la même caractéristique, d’où l’intérêt porté aux langues austronésiennes, Niger-Congo et Nilo-sahariennes, turciques, slaves et romanes. Dans la mesure où nous avons décidé de travailler sur les constructions antipassives dérivées par une marque étant à l’origine polysémique, l’impact sémantique de cette dernière sur l’ensemble de la construction apparaît comme non négligeable. Cette étude présente ainsi le double intérêt de s’appuyer sur une approche translinguistique, impliquant différentes familles de langues, et sur une vision bipolaire relative aux domaines de la syntaxe et de la sémantique. / The antipassive term, associated with ergative system since forty years, denotes an intransitive construction in which the patient argument is syntactically demoted. This study calls into question a traditional opinion according to which the antipassive phenomenon is encountered in ergative languages but not in those of accusative alignment. Since in some ergative languages the antipassive construction is triggered by a polysemous reflexive and/or reciprocal morpheme, this study deals exclusively with those accusative languages in which the antipassive marker presents the same characteristics (cf. Austronesian, Niger–Congo, Nilo-Saharan, Turkic, Slavonic, Romance languages). Building on the polysemous nature of such marker, its possible semantic impact on the whole derivation should also be taken into consideration. The aim of this study is to analyse the antipassive construction both from semantic and syntactic point of view in a crosslinguistic perspective.

Argument coding splits such as differential (= split) object marking and split ergative marking have long been known to be universal tendencies, but the generalizations have not been formulated in their full generality before. In particular, ditransitive constructions have rarely been taken into account, and scenario splits have often been treated separately. Here I argue that all these patterns can be understood in terms of the usual association of role rank (highly ranked A and R, low-ranked P and T) and referential prominence (locuphoric person, animacy, definiteness, etc.). At the most general level, the role-reference association universal says that deviations from usual associations of role rank and referential prominence tend to be coded by longer grammatical forms. In other words, A and R tend to be referentially prominent in language use,while P and T are less prominent, and when less usual associations need to be expressed, languages often require special coding by means of additional flags (casemarkers and adpositions) or additional verbal voice coding (e.g., inverse or passive markers). I argue that role-reference associations are an instance of the even more general pattern of form-frequency correspondences, and that the resulting coding asymmetries can all be explained by frequency-based predictability and coding efficiency.

No description available.