Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Janet McDonald


A series of experiments explored the effects of performing concurrent secondary tasks on learning letter strings created with a finite state artificial grammar. Experiments 1-2 compared a task which disrupted organized encoding to a task which simply required holding information in memory while encoding strings, as well as to two control tasks. Participants performing the disruptive task were worse at judging the grammaticality of test strings than were participants in the two single task control groups. Performance of the memory load group fell between the disruptive task group and the control groups, but was not significantly different from either. Experiments 3-4 compared the effects of secondary tasks which consistently grouped letters frequently seen together or consistently interrupted letters frequently seen together in grammatical strings. Disrupting frequent letter groups inhibited learning to a greater extent than grouping frequent chunks; however, predicted facilitatory effects for chunking frequent groups of letters were not found. Experiment 4 also tested the effects of secondary task stimuli differing on relative verbalizability, finding very little difference amongst the three types of stimuli tested. Additionally in Experiments 2 and 4, the ability to detect ungrammatical strings with violations in various locations was tested. These results replicated previous findings in the grammar learning literature, with errors at the beginnings and ends of strings easier to detect than those in the middle. Findings of this research indicate that it is possible to learn artificial grammar strings under dual task conditions; however, performing any type of secondary task is likely to inhibit learning somewhat. The extent of disruption may depend on the processing demands of the secondary task. Overall results indicate that a full explanation of the effects of secondary tasks on grammar learning may require a two factor model including both limited capacity processing resources and necessary organization of study strings.