2009 Fellowship Recipients

The citizens of Afghanistan have endure extraordinary hardships over the past decade, yet the identification of specific at risk populations and areas that have been hindered by a lack of detailed and reliable information. This project explores the limitations and possibilities of three national household surveys conducted in Afghanistan (2006-08) by the Asia Foundation, funded through USAID, as a potential source of information concerning household health and well-being. Dr. Buckley plans statistical analyses focusing on the social and economic correlates of household health decline, prioritization of health care quality, perceptions of healthcare corruption and reported healthcare access in Afghanistan. These analyses will inform our understanding of population health and attitudes towards healthcare, as well as their variation across the region, household composition, age, gender and socioeconomic status. As health operates as an especially valid indicator of social standing in times of acute crisis, this work will contribute to overall discussions of stratification and social need in Afghanistan, as well as theoretical discussions focusing on the influence of social conflict and disorder on health and well-being. Dr. Buckley intends to use these analyses to guide the generation of a larger, multi-method, and collaborative research proposal focusing on the intersections of gender, health and household economic stability in Afghanistan.

Proposed here is a project to develop a prototype computer program that depicts the meaning of an input stream of natural-language textual material.  The depiction appears on a graphical interface and changes in real time as the program recognizes and processes each sequential morpheme (minimally meaningful unit of speech) in the input stream.  Video-game quality of the graphical depiction is achieved using an appropriate video-game engine embedded in the application program.

The proposed project targets two languages of Afghanistan, Kâmviri (a language of eastern Nurestân Province) and Pashto, as languages to be depicted by the computer program.  The proposed project expands on the applicant’s previous and on-going linguistic and computer-program-development work (Strand 1985, 1991, 1999-2000, 2006), by increasing the verisimilitude of depiction through the use of video-game software technology and by increasing the number of morphemes in Kâmviri and Pashto that can be recognized and depicted.

The recognition of morphemes and subsequent depictional processing are controlled by a plug-in dictionary program-module, which must be designed individually for each language that is input.  The proposed project will accordingly be implemented in two concurrent phases: development of the computer program and development of dictionary modules for the two targeted languages.

As of May 2009, the Speech Depicter program will have the capability to depict the actions of simple sentences in the target language, properly handling and depicting such grammatical processes as pronoun reference, verbal aspect and mode, prepositional and postpositional specification of spatial relationships among objects, and adverbial specification of the temporal relationships of events. The Speech Depicter program has multi-user, multilingual capability, so that users speaking any of the target languages can communicate through a common visual scene. The project’s second phase will produce a crucial increase in the verisimilitude of the depiction, by depicting more grammatical forms and by expanding the dictionary modules for each of the targeted languages to depict more types of verbs and more kinds of motion for human characters participating in verbal action.