Stuck on semantics: Processing of irrelevant object-scene inconsistencies modulates ongoing gaze behavior.

Area of Science:

• Cognitive Psychology
• Visual Perception
• Neuroscience

Background:

• Humans possess remarkable abilities for rapid object and scene identification.
• The automaticity of scene and object semantic processing, especially semantic congruity, remains an active area of research.
• Previous research suggests that objects incongruent with their scene context may elicit different visual attention patterns.

Purpose of the Study:

• To investigate whether semantic congruity between objects and scenes modulates eye movement behavior, even when irrelevant to the primary task.
• To determine if attention is automatically allocated to semantically incongruent objects.
• To explore the impact of semantic incongruity on memory recall for objects within a scene.

Main Methods:

• Participants performed a visual search task for a target letter ('T') overlaid on indoor scene photographs.
• Some scenes contained objects that were semantically incongruent with the overall scene context (e.g., a toothbrush in an office).
• Eye-tracking was used to measure gaze behavior, and subsequent memory tests assessed explicit and implicit recall.

Main Results:

• Participants spent significantly more time looking at semantically incongruent objects compared to congruent objects, despite the objects being irrelevant to the search task.
• Memory tests revealed limited recall for both congruent and incongruent objects, suggesting that prolonged fixation does not necessarily equate to robust memory encoding.
• Eye movement behavior was modulated by semantic mismatches, indicating obligatory processing of scene-object relationships.

Conclusions:

• Scene and object relationships are processed obligatorily during naturalistic viewing.
• Irrelevant semantic mismatches between scene and object identity can automatically influence and modulate ongoing eye-movement behavior.
• This automatic processing occurs even when the semantic information is not relevant to the participant's current task.