List of Past Events
Alternative Minimalist Visions of Language
Dr. Ray Jackendoff
Tuesday, October 18, 2005, 01:00pm - 02:00pm
The primary goal of linguistic theory is to explain how the child attains adult competence in language. A theory satisfying this goal must meet two constraints. First, it must be descriptively adequate at accounting for adult competence. Second, it should attempt to minimize what had to evolve in order to get the innate human faculty of language from the ape genome. The Minimalist Program addresses these constraints by positing that the grammar has an absolute minimum of machinery; complexity arises only by virtue of interactions with independent properties of sound and meaning. Thus there is little for the child to learn. An alternative minimalist strategy is to find ways to formulate the complexity of adult grammar in such a way that more of it can be learned. Part of this strategy is to minimize elements of linguistic structure that cannot be inferred from overt form, such as null elements, covert syntactic structure (e.g. LF), and movement that relates covert syntax into surface form. Such is the strategy available in constraint-based monostratal theories such as HPSG and Construction Grammar. The choice between these two approaches to minimalism is an empirical one. It will be argued that minimizing structure rather than rules is the correct strategy. At the same time, it will be shown that two fundamental tenets of mainstream generative grammar since its inception must be replaced. First is the assumption that the formalism for describing the grammar involves a derivation, a logical sequence of manipulations that build and/or restructure trees. This will be replaced with a constraint-based formalism, common to a number of theoretical frameworks, in which all rules apply simultaneously to license the structure of trees. A second defective assumption of mainstream generative grammar, in this case inherited from traditional grammar, is that there is a strict distinction between the rules of grammar and the lexicon. This will be replaced with the view, again arrived at within a number of frameworks, that there is no strict distinction between lexical items and rules, but rather a cline of specificity vs. generality. It will be shown that this approach leads to a potentially tractable account of rule learning. The overall prediction: If any approach to language is going to make meaningful contact with cognitive neuroscience and evolutionary biology/psychology, it will be an approach growing out of constraint- and construction-based minimalism, not out of the Minimalist Program.