Submitted theses 2016

Thesis: The Ngkolmpu Language with special reference to distributed exponence

Matthew Carroll

PhD: Matthew Carroll

Submission year: 2016

Supervisors: Nicholas Evans & I Wayan Arka

CoEDL node: Australian National University 

CoEDL program: Shape

Abstract:

The Ngkolmpu language of southern New Guinea is notable for the remarkable extent to which grammatical values are distributed across multiple morphosyntactic systems in the language. This is most apparent in the extremely complex inflectional morphology of verbs, where the exponence of morphosyntactic feature values is distributed over a number of inflectional sites, such that determining the exact value of any given feature requires unification at multiple structural locations. Moreover, this phenomenon is not restricted to the inflectional morphology, and permeates the morphology, syntax and semantics of the language. This thesis provides the first comprehensive description of the phonology, morphology and nominal and clausal syntax of Ngkolmpu.

 
 

Thesis: The role of the native language in non-native perception and spoken word recognition: English vs. Spanish learners of Portuguese

Jaydene Elvin

PhD: Jaydene Elvin

Submission year: 2016

Supervisor: Anne Cutler

CoEDL node: Western Sydney University

CoEDL program: Processing

Abstract:

The ultimate goal for adult learners of a second language (L2) is successful communication. If learners cannot perceive, recognise and produce sounds and words in the L2 they may struggle to understand speakers of that language, who may in turn struggle to understand L2 learners. Not all learners will attain the same level of proficiency and even when immersed in the L2 environment, difficulties in L2 speech perception, spoken word recognition and L2 production persist. These difficulties in L2 speech are often attributed to the influence of the native language on the acquisition of L2 speech perception, spoken word recognition and production. This thesis investigates the role of the native language in Australian English (AusE) and Iberian Spanish (IS) listeners’ non-native vowel perception and spoken word recognition of Brazilian Portuguese (BP) and the interrelation between these two abilities. The thesis also investigates whether or not individual listeners follow different developmental patterns when perceiving and recognising BP. 

Open access: https://researchdirect.westernsydney.edu.au/islandora/object/uws%3A37801

 

Thesis: Evolving spatial and temporal lexicons across different cognitive architectures

Scott Heath

PhD: Scott Heath

Submission year: 2016

Supervisors: Janet Wiles & David Ball

CoEDL node: University of Queensland

CoEDL program: Technology

Abstract:

Communication between mobile robots requires a transfer of symbols, where each symbol signifies a meaning. However, in typical applications, meaning has been ascribed to the symbols by the engineers that have programmed the robots. This thesis explores an alternative: the use of algorithms and representations that allow mobile robots to evolve a shared set of symbols where the meanings of the symbols are derived from the robots' sensors and cognition.

Open access: https://espace.library.uq.edu.au/view/UQ:377716

 

Thesis: Distributional Learning of Lexical Tone and Musical Pitch by Naive and Experienced Adult Learners

Jia Hoong Ong

PhD: Jia Hoong Ong

Submission year: 2016

Supervisor: Anne Cutler

CoEDL node: Western Sydney University 

CoEDL program: Processing

Abstract:

Language and music are two human universals that share many commonalities, including processes of statistical and distributional learning in acquiring knowledge of those domains. This thesis is concerned with the role of distributional learning in the acquisition of pitch-based building blocks of speech and music. In a series of five studies, questions of theoretical and empirical interest will be examined, whether: (i) distributional learning can be used to acquire lexical tone and musical pitch; (ii) domain-general or domain-specific pitch experience facilitates distributional learning of pitch; and (iii) distributional learning plays a role in cross-domain transfer. The results of all five studies suggest that distributional learning can be used to acquire the foundations of speech and music; using distributional learning, adult learners either shift existing category boundaries to which the perceptual items assimilate or form new categories if the perceptual items are not assimilated to any native (linguistic or musical) categories. While distributional learning appears to be sensitive to top-down interferences and is modulated by domain-specific experience, it is nonetheless a powerful learning mechanism that is generalisable across domain. This thesis thus advances our understanding of speech and music by providing evidence for the commonality between the two in terms of a common learning mechanism and shared pitch processing, both of which are compatible with accounts of a common origin for language and music.

Open access: https://researchdirect.westernsydney.edu.au/islandora/object/uws%3A37579

  • Australian Government
  • The University of Queensland
  • Australian National University
  • The University of Melbourne
  • Western Sydney University