Report on Grant Activities

We recently submitted the first annual report to the National Science Foundation regarding activities conducted on the grant. An abridged version of the report is given here.

The overarching goal of the project is to reevaluate and expand what we know about the typology and history of Australian languages, and to test new methods for investigating the past. Progress has been made in all areas over the last 12 months. Work focuses on two general research questions:

  • basic research in Australian languages, including language documentation from existing materials
  • investigating process of language change, to refine theories of how languages change over time.

The project narrative proposed four strands of investigation into these questions:

  1. documentation: lexical data entry, boot camp, language identification, data dissemination
  2. typology: investigation of phonological traits
  3. reconstruction
  4. phylogenetics

Discussion of the four strands of investigation follows.

1. Documentation

  • The REU ‘grammar boot camp’ is currently underway. Three students have been writing a grammar of the Ngalia language of southwestern Australia, using materials provided by Susan Hanson and Kado Muir from the Goldfields Aboriginal Language Centre. The book is currently about 130 pages and is on track for submission to Pacific Linguistics’ A-PL online grammar series in August.
  • Lexical data entry to the comparative database continued, but it has not been a high priority at present. Rather, we have been cleaning existing data in preparation for release of the first stage of the database, most likely in August, 2015. We have added approximately 5,000 new items and sourced about 10 new lexicons for inclusion in a subsequent release.
  • Also in preparation for data dissemination, a student researcher has contacted stakeholders about data to secure up to date information about access restrictions (if any). Almost everyone who has been contacted has been keen for their materials to be generally released. Two other researchers required speech community consultation before releasing the materials.
  • Bowern has been in discussions with Yale’s Instructional Technology Services to arrange release of the database files.
  • Joshua Phillips and Carmen Wilson researched the use of off-the-shelf text-to-speech alginment software for Australian languages. They completed the manual measurements for evaluating the results. Write-up is in progress.
  • Rikker Dockum has been working on ways to associate existing lexical glosses with a set of standardized gloss forms, to facilitate searching in the database. He has also compiled a list of materials to add to the database which could be done with OCR.

2. Typology

  • A student has been extending earlier work by grant personnel (Bowern and Gasser 2014) on phonemic typologies of Australian languages. This has involved cleaning and expanding the data set, generating hypotheses about segment distribution, writing python scripts to extract the relevant data, running those scripts, and analyzing the results.
  • Our collaborators at the University of Queensland have been researching how segmental inventory typologies and segment transition probabilities could be used in phylogenetic analyses. An honours thesis (= senior thesis) on this topic was recently submitted.

3. Reconstruction

  • Two students worked on a typology of polysemy relations, as part of reconstruction of specific semantic domains. They replicated William Croft’s work on Dravidian and Indo-European colexification patterns and wrote
  • Bowern has been commissioned to write an article on linguistic isolates in Australia.
  • Zhou and Haynie have been involved in developing and implementing computational methods for long-range lexical comparison. This has involved the evaluation of metrics for assessing linguistic relatedness in an Oswalt Monte Carlo framework, and helping to supervise student researcher Kevin Zhou’s implementation of an iterative alignment program for the comparison of lexical data.
  • Color term phylogenetics: This project uses Bayesian phylogenetic methods to model the evolution of color term systems in Pama-Nyungan languages, a large language family whose daughter languages employ diverse color naming systems. This project involves both the reconstruction of ancestral color systems and also the evaluation of proposed pathways of color term evolution (e.g. Berlin and Kay 1969). Work on this project has included coding of large multi-state datasets, implementing models in BayesTraits software, applying traditional comparative methods to individual color terms within Pama-Nyungan subgroups, and additional analysis of model results.

4. Phylogenetics

  • Haynie has been mostly working in this area. She participated in a workshop “towards a global language phylogeny.” Sept. 17-19, 2014. Max Planck Institute for the Science of Human History. Jena, Germany. She also participated in a GlottoBank Workshop. Feb. 23-26. University of Auckland and Max Planck Institute for the Science of Human History. Waiheke, New Zealand.
  • Ongoing participation in the GramBank component of the MPI Jena GlottoBank project: Haynie has been involved in the design of a questionnaire for collecting a standard set of grammatical features targeted for use in phylogenetic and comparative linguistic research. The aim of the project is to create a global database of structural linguistic features to serve as an empirical basis for research on deep historical relationships between languages. She is also involved in the LexiBank component of the GlottoBank project, which is focused on word list data, lexical comparison, and sound change.
  • GIS analysis of linguistic data:  Haynie has also been involved in several projects aimed at using geographic analysis to identify spatial patterns associated with the diffusion of linguistic material, both at the language area scale (e.g. by looking at areality in phonological systems) and at smaller scales (e.g. through collaborative work with the Yale Grammatical Diversity Project). Though she is currently exploring the use of GIS tools for identifying and examining linguistic areas in convenient American datasets, these methods being piloted are relevant to questions regarding the role of diffusion in Australian language relationships.

Comment on this post