Integration versus interactive activation: the joint influence of stimulus and context in perception

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Linguistics

Background:

• Empirical data show independent influences of stimulus and context on perceptual recognition.
• The Fuzzy Logical Model of Perception (FLMP) and Interactive Activation and Competition (IAC) models offer different explanations.
• Previous IAC models struggled to account for independent influences.

Purpose of the Study:

• To compare the predictive power of the FLMP and modified interactive activation models (SIAC, Boltzmann machine) for perceptual recognition.
• To evaluate how these models account for the dynamics of information processing over time.
• To determine which model better describes the joint influence of stimulus and context in language perception.

Main Methods:

• Formulated dynamic versions of the SIAC and FLMP models to predict performance as a function of processing time.
• Tested quantitative predictions against empirical results from reading and speech perception.
• Analyzed the models' ability to account for context effects and stimulus influence at different processing times.

Main Results:

• The dynamic FLMP provided a superior description of the time course of perceptual processing compared to interactive activation models.
• SIAC and Boltzmann models showed limitations in predicting context effects with limited processing time and stimulus influence with substantial processing time.
• FLMP accurately predicted that context can enhance performance accuracy and introduce response bias.

Conclusions:

• Empirical and theoretical evidence favors the FLMP over SIAC models for pattern recognition.
• Interactive activation models are less consistent with empirical results and not necessary for explaining joint stimulus-context influences in perception.
• The FLMP offers a more robust framework for understanding perceptual recognition dynamics.