Lexicalization in adults who stutter

Three recent theories have implicated lexical processing failures as a possible source of fluency disruption in persons who stutter. Two experiments that bear upon these theories are reported. Both evaluate the effects on speech response latency of picture naming tasks designed to place selective demands on lexicalization: Experiment I, effects of one-word versus two-word responses; Experiment II, effects of a word's frequency of occurrence versus its number of syllables. Twelve adults who stutter and 12 with normally fluent speech participated in each experiment. In Experiment I, increases in naming latency for two-word (noun + verb) versus one-word (noun or verb) responses showed that demands for parallel processing did not differentiate the experimental groups. However, the between-group difference, showing longer latencies among those who stutter, was six times greater for the verb, than for the noun, task. Moreover, the group difference for verbs fully accounted for the size of the difference in the two-word task. Experiment II showed that the relative increase in naming latency associated with the word frequency effect versus the syllable latency effect was significantly greater in the stuttering than the nonstuttering group. Outcomes of the two experiments suggest that during lexicalization, as early as the L1 stage and first phase of L2, slow processing could serve to disrupt fluency in some persons who stutter. Under certain conditions, as specified in the three theories cited, such disruptions could set the occasion for stutter events.