The content of the phonology seminar changes with each teaching.
Fall 2008
This seminar will focus on four broad themes.
1. "Channel" vs. "analytic" biases
An important area of debate these days centers on the extent to which phonological markedness follows from something like "competence" in the generative sense vs. "accidents" of transmission (the latter as posited by Evolutionary Phonology). The term "competence" is so laden with past interpretations as to be almost unhelpful, and we might prefer Moreton's terms: "channel bias" for distortions introduced by language transmission (as when misperception occurs); "analytic bias" for cognitive predispositions toward some patterns over others. There is a growing body of research that seeks to determine by experimental means which biases, if any, are analytic. In such experiments, subjects are presented with artificial language data, the goal being to see how the subjects are constrained in learning or generalizing patterns. When these constraints cannot be argued to follow from the subjects' own native language, we can (arguably) infer an analytic bias. In this portion of the course we will read and discuss several papers in this area and try to understand the experimental methodologies and what we think they really tell us.
2. Contrast dispersion as emergent?
By its very nature the notion of contrast dispersion seems to presuppose that speakers compare forms with other contrasting forms. One might wonder whether this is really true, and if it is, how it happens. For people like me who have practiced dispersion theory in an OT guise, the question becomes one of how we determine the "comparison set" of forms for the purposes of constructing OT tableaus. In this portion of the course we will explore an alternative to the OT approach. This alternative is based on viewing contrast dispersion as an emergent property, as argued for by e.g., Wedel, or Boersma and Hamann, and it amounts to an Evolutionary Phonology approach to dispersion.
3. Computational/mathematical modeling of contrast dispersion
Since the seminal work of Björn Lindblom we have seen various attempts to predict the typology of vowels, using mathematical or computational simulation, based on what is known about vowel perception. A well known recent model, called Dispersion-Focalization Theory (DFT), is due to Schwartz et al. (1997). In recent work, Nathan Sanders and I have been exploring the properties of DFT in some depth, and seeking ways of improving its predictions. This portion of the course discusses our progress in this area.
4. Derivational opacity and russian phonology
A recent contribution to the thorny problem of derivational opacity is McCarthy's candidate chain theory (OT-CC). We will read McCarthy's work in this area. As a way of trying to understand it, we will apply it to problems of Russian phonology. This is selfish of me, because I like Russian phonology.
Return to Jaye Padgett's Teaching.
Last modified December 23, 2008.