Planning, language, and comprehension: Knowledge-based inferences are NOT general

Comprehension, which relies on the inference process (Graesser et al., 1994), depends on general knowledge to support causal explanations between text ideas. Processing differences between two areas of knowledge (physical cause-effect vs goals/planning) were explored in order to establish (1) that different areas of knowledge would similarly support inferences, and (2) that differences in knowledge areas, but not inference abilities, would impact comprehension after brain injury (ABI). ABI participants who had impaired planning skills were compared to gender, age, and education level matched non-injured participants (NI). Experiment 1 examined the hypothesis that participants with ABI would make physical, but not planning, inferences. NI participants were expected to make inferences from both knowledge areas. Participants read two-sentence texts describing a causal relation (inference) or not (control) and provided measures of inference processes by probe recognition and comprehension questions. NI participants demonstrated equal inferences across knowledge areas, indicating both knowledge areas supported the inference process. ABI participants demonstrated inferences primarily for physical knowledge, indicating the inference process was intact, but planning knowledge was less available to support comprehension. This result indicated the separability of these knowledge areas.; Experiment 2 utilized a dual-task paradigm to consider whether differences between knowledge areas involved resource demands on working memory and attention in NI participants. Reading and comprehension of the same two-sentence texts from experiment 1 constituted the primary task. The secondary task manipulated resource load by requiring participants to simultaneously rehearse either a low load (three words) or high load (six words) in working memory. The dual-task manipulation resulted in a difference between knowledge areas for the probe recognition measure, but not the comprehension question measure. Probe recognition measures of inference processing were maintained under low load for both knowledge areas. While under high load, evidence of inference processes were maintained from physical knowledge but eliminated for planning knowledge. These findings suggest there may be a resource requirement difference for these two areas of knowledge. Combined, these data suggest areas of general knowledge may function differentially to support the inference process involved in comprehension.