Spatial Effects on Semantic Competition in Discourse Comprehension 1
2
Glenn Williams , Anuenue Kukona , & Yuki Kamide
1
[email protected] 1 School of Psychology: University of Dundee, Dundee, UK 2 School of Applied Social Sciences: De Montfort University, Leicester, UK
Results Continued
Introduction Ongoing events are organised into separate event models around distinct locations (Radvansky, 2012). Only relevant models (and information in them) are accessed (Ezzyat & Davachi, 2011) during comprehension.
Looks to each object when together (left) and apart (right)
Previously, we described movement of a book, causing the old representation (on the floor) and the updated representation (on the table) to be (a) together or (b) apart in separate models (i.e. rooms):
B. Apart
A. Together
During “lock is”... 1.The woman will take the book to the table. 2.Then she will study the painting and pick up the book. During book in 2, fixations on the table: Together > Apart Separating the target and (same object) competitor necessitates activating 2 relevant models (rooms) increasing interference
Together: LOCK > KEY > BALL = MELON Apart: LOCK > KEY = BALL = MELON
Looks to the target (left) and competitor (right) when together vs. apart
If the competitor is a different (irrelevant) object (e.g. semantic interference; fixating on KEY when hearing LOCK (Yee & Sedivy, 2006)), does separation reduce interference? Hypothesis: Semantic interference is reduced by associating the LOCK and KEY with separate event models
Methodology Sixty participants viewed 32 experimental quadrants with a target (LOCK) competitor (KEY) and two distractors (BALL, MELON) Concurrent visual world paradigm, ‘look and listen’ task Critical manipulation: location of LOCK relative to KEY — together (1a, b) or apart (1c, d)
During “lock is”... Fixations on the LOCK: Together < Apart Fixations on the KEY: Together > Apart
1a. The lock and the key are in the cafeteria. The ball and the melon are in the parlour. 1b. The ball and the melon are in the cafeteria. The lock and the key are in the parlour. 1c. The lock and the melon are in the cafeteria. The ball and the key are in the parlour. 1d. The ball and the key are in the cafeteria. The lock and the melon are in the parlour.
Event models guide fixations on and influence accessibility for (i) target objects, and (ii) semantic competitors
2. It seems that the lock is very old.
Event interference: Separate event models are only accessed when the competitor is relevant to the target (i.e. same object competitor)
Predictions: During “...lock is...”
Together (1a,b): LOCK > KEY > BALL = MELON Apart (1c, d): LOCK > KEY = BALL = MELON LOCK: Together < Apart; KEY: Together > Apart
Results Hierarchical mixed effects models with an empirical logit transformation (Barr, 2008) were conducted during the “lock is” region. Separate analyses were carried out for subjects and items; reported findings are significant by both.
Discussion Semantic competition is observed when the target and competitor are together and disappears when apart
Object Interference: Only the different object competitor (vs. same object) is fixated on more often; not suppressed by updating Future study: Exploring whether visual competition between similarly shaped objects (e.g. baseball bat, cigarette) can be modulated by event structure. Here, other sources of visual similarity (e.g. colour) have been removed. This experiment is currently underway.
Barr, D. (2008). Analysing ‘visual-world’ eyetracking data using multilevel logistic regression. Journal of Memory and Language, 59 (4), 457-474.
Radvansky, G. A. (2012). Across the event horizon. Current Directions in Psychological Science, 21 (4), 269-272.
Ezzyat, Y., & Davachi, L. (2011). What constitutes an episode in episodic memory? Psychological Science , 22 (2), 243–252.
Yee, E., & Sedivy, J. C. (2006). Eye movements to pictures reveal transient semantic activation during spoken word recognition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32 (1), 1–14.
