Causal inferences in reading: from immediate activation to long-term memory

Area of Science:

• Cognitive Psychology
• Psycholinguistics
• Memory Studies

Background:

• Understanding causal inference is key to comprehending text.
• The temporal dynamics of how readers form these inferences remain incompletely understood.
• Reactivation of prior concepts is hypothesized to play a role in resolving coherence breaks.

Purpose of the Study:

• To investigate the time course of drawing causal inferences during reading.
• To determine if readers reactivate earlier concepts to resolve causal coherence breaks.
• To examine the retention of these inferences in working memory and long-term memory.

Main Methods:

• Three experiments utilized reading comprehension tasks with causal coherence breaks.
• Experiment 1 employed a probe word naming task to assess concept reactivation.
• Experiment 2 used reading times to measure inference encoding in working memory.
• Experiment 3 utilized a recall task to evaluate long-term memory representation.

Main Results:

• Participants demonstrated faster naming of probe words related to the causal concept after a coherence break.
• Reading times increased significantly for sentences contradicting the inferred causal link.
• Recall tasks indicated that the causal link was successfully integrated into long-term memory.

Conclusions:

• Causal inferences are drawn rapidly, involving quick reactivation of relevant concepts.
• Inferences are actively maintained in working memory, influencing subsequent processing.
• The study provides evidence for the integration of causal links into long-term memory representations of text.