This syntax needs learned: Adult acquisition of novel syntactic constructions

Many current psycholinguistic theories are built on models of learning and memory. As such, they simultaneously serve as theories of language acquisition and language processing. This dissertation evaluated three types of theories (constraint satisfaction accounts of sentence processing, theories of structural persistence, and episodic-processing accounts of memory) with regard to their adequacy in accounting for a phenomenon that has been documented in the field studies of sociolinguists, but rarely studied in an experimental setting: adults' ability to learn new syntactic constructions in their native language. Adults were trained on a novel construction called the Needs construction (e.g., “The meal needs cooked”). Experiment 1 used a text-reading paradigm to demonstrate that adults quickly learn to comprehend the Needs construction. Experiment 2 replicated this finding in a word-by-word reading task, and showed that adults readily generalize the new construction to a verb similar to “needs” (e.g., “wants”). Experiments 3–5 assessed the effect that learning the Needs construction has on the processing of sentence structures that already exist in the participants' language. These experiments explored how learning the Needs construction affects processing sentences such as, “The meal needs cooked vegetables to make it complete” (i.e., the Modifier construction). Experiment 3 showed that participants read the Modifier construction faster when they have been trained on the Needs construction than when they have been trained on the version of the Needs construction that is standard in their dialect (e.g., “The meal needs to be cooked”). This finding is consistent with episodic-processing memory theories, but not with constraint satisfaction theories and theories of structural persistence. Experiment 4 demonstrated that this facilitation is attenuated when the Modifier construction is altered so that it does not resemble the surface form of the Needs construction as closely. Experiment 5 provided further support for the episodic-processing account. This experiment showed that the effect of Needs construction training on the reading of Modifier sentences depends on the nature of the experience that learners have during their training. The implications of these findings for psycholinguistic theory, the methodology of psycholinguistic experiments, and language change are discussed.