Bruce Tesar & Paul Smolensky (1998): "Learnability in Optimality Theory." Linguistic Inquiry 29: 229–268.
The first version of the constraint-demotion algorithm appeared in:
Bruce Tesar & Paul Smolensky (1993): "The learnability of Optimality Theory: an algorithm and some basic complexity results." Manuscript Department of Computer Science & Institute of Cognitive Science, University of Colorado at Boulder. Available as Rutgers Optimality Archive 2, http://ruccs.rutgers.edu/roa.html
The Error-Driven Constraint Demotion algorithm (EDCD) can be found in:
Bruce Tesar & Paul Smolensky (1996): "Learnability in Optimality Theory (long version)". Technical Report 96-3, Department of Cognitive Science, Johns Hopkins University, Baltimore. Available as Rutgers Optimality Archive 156, http://ruccs.rutgers.edu/roa.html
A corrected version of EDCD appeared in Boersma (2009b).
© ppgb, March 31, 2010