Testing the Altercentrism Hypothesis in Young Infants

• In Event: Theory of mind development and children’s representation of beliefs

Abstract

A series of conditions with 7.5-to-8.5-months-old infants were designed to test the altercentrism hypothesis (Southgate, 2020). The hypothesis has two parts relevant to development. First, it proposes that infants use others as ‘learning devices’: relatively immobile infants spend a large proportion of their time watching their caretakers interacting with the environment. Others’ attention and actions drive infants’ attention to relevant objects, and enables preferential encoding of properties or relations relevant for others. Second, infants are proposed to rely most heavily on this device, encoding relatively less of objects or events that were not the targets of others’ attention.
We tested this hypothesis with location-memory for an object behind one of two occluders, and asked whether 7-month-old infants represent a co-witnessed occlusion event better than a subsequent displacement witnessed alone. Our crucial condition follows the logic of False Belief (FB) tasks (Onishi & Baillargeon, 2005), with the difference that the agent never returns to search at the end. Instead, we reveal the object absent from behind the first occluder (co-witnessed with the agent) or absent from the second, where it should be (but only the infant has seen that). We contrast this condition with a True Belief (TB) manipulation, where the agent is present only for displacement of the last location. Again, we alternately reveal the two empty locations over different trials and we measure looking time and first look duration.
In addition, two baseline conditions establish (1) whether our paradigm can reveal memory for location and (2) whether infants remember object locations better when their occlusion is brought about by the actions of an agent. In one the object was placed in the two locations by a hand (agency cue) and the other condition was a non-social setting where a conveyor belt transports the object.
The altercentric hypothesis predicts that the first (co-witnessed) displacement will be better remembered in the FB than in the TB condition and that we find a stronger effect for the second displacement in TB than in baselines.
To date, data collection is completed for the two baseline conditions (n=32 each) and data collection is at n=16/32 for FB and n15/32 for TB. We expect to finish testing in fall 2020.
We found evidence for location memory with one of our preregistered measures, first look (F = 5.526, p = 0.022) but not with total looking time (F = 3.061, p = 0.087). And we found no effect of condition when comparing conveyor belt vs. hand. An exploratory Bayes Factor reveals moderate evidence for the null, condition-wise (6.97 for first look, 3.56 for looking time). Final results by fall 2020.

Authors

• Velisar Manea, University of Copenhagen
  Presenting Author

• Dora Kampis, University of Copenhagen
  Non-Presenting Author

• Charlotte Grosse Wiesmann, Max Planck Institute for Human Cognitive and Brain Sciences
  Non-Presenting Author

• Victoria Southgate, University of Copenhagen
  Non-Presenting Author