Dynamics of activation in semantic and episodic memory.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Computational Linguistics

Background:

• Spreading-activation models describe semantic and episodic memory as interconnected networks.
• A key debate concerns whether activation growth at network nodes is discrete or continuous.
• Understanding activation dynamics is crucial for explaining memory processes like word recognition and comprehension.

Purpose of the Study:

• To empirically test the dichotomy between discrete and continuous activation growth models in memory networks.
• To differentiate between abrupt, state-based activation changes and gradual, continuous activation increases.
• To assess the validity of two-state discrete activation models against continuous models.

Main Methods:

• Utilized an adaptive priming procedure involving prime and test stimuli with varying intervals.
• Subjects performed lexical decisions on test stimuli, with prime-test intervals adjusted based on performance.
• Analyzed reaction time distributions as a function of prime-test interval duration to identify activation patterns.

Main Results:

• The predicted finite mixture distribution for discrete activation models did not emerge in experiments.
• Results consistently failed to support discrete activation models in both semantic and episodic memory tasks.
• Two-state discrete activation models were rejected based on the experimental data.

Conclusions:

• The findings reject discrete activation models for semantic and episodic memory.
• Evidence suggests that activation growth in memory networks does not follow abrupt, distinct states.
• Further research is needed to refine continuous activation models or explore alternative explanations.