A New View on Language and Cognition
Data-Oriented Parsing (DOP) models of natural language propose that human language perception and production work with representations of concrete language experiences, rather than with abstract grammatical rules. These models operate by decomposing the given representations into fragments and recomposing those pieces to analyze new utterances. DOP is therefore a new way to think of what a native speaker knows when s/he knows a language.
DOP framework has been applied to other modalities as well, such as music,
vision, problem-solving and reasoning. DOP is therefore a candidate to solve Allen Newell`s
challenge, the search for a single model that unifies perceptual and
cognitive modalities. The VICI
Group at the
Here is a light-hearted view on DOP
An accessible introduction to DOP is a tutorial paper by R. Bod and R. Scha, 1996.
A short tutorial into DOP is Sima`an 2003.
For a state-of-the-art overview, see the DOP edited volume.
An overview of DOP applied to other modalities is here.