Infants’ understanding of the causal power of agents and tools

post by Bruce W. Lee (bruce-lee) · 2024-02-27T18:36:42.037Z · LW · GW · 0 comments

This is a link post for https://www.pnas.org/doi/10.1073/pnas.2309669120#:~:text=The%20present%20studies%20explored%20how,the%20laws%20of%20intuitive%20physics.

Contents

  Significance
  Abstract
  Overview
  Empirical
    Experiment 1
    Experiment 2 (Control)
    Experiment 3
    Experiment 4
  Ultimatum
None
No comments

Significance

Tool-use understanding is classically thought to involve reasoning about the causal mechanisms that relate tools and their effects (e.g., moving food closer). Through cultural evolution and accumulated innovations, however, humans have developed tools that became causally opaque (e.g., touch screens). We show that cultural transmission remains possible because even young infants are equipped with the fundamental capacity to represent arbitrary causal events where an agent acts on the world with a tool. Infants attributed to a ball have the power to cause an arbitrary state change but only if that ball was itself launched by an agent. Infants thus consider that the power to cause arbitrary events ultimately belongs to intentional agents but agents can transfer this power to tools.

Abstract

Tools are objects that are manipulated by agents with the intention to cause an effect in the world. We show that the cognitive capacity to understand tools is present in young infants, even if these tools produce arbitrary, causally opaque effects. In experiments 1–2, we used pupillometry to show that 8-mo-old infants infer an invisible causal contact to account for the—otherwise unexplained—motion of a ball. In experiments 3, we probed 8-mo-old infants’ account of a state change event (flickering of a cube) that lies outside of the explanatory power of intuitive physics. Infants repeatedly watched an intentional agent launch a ball behind an occluder. After a short delay, a cube, positioned at the other end of the occluder began flickering. Rare unoccluded events served to probe infants’ representation of what happened behind the occluder. Infants exhibited larger pupil dilation, signaling more surprise, when the ball stopped before touching the cube, than when it contacted the cube, suggesting that infants inferred that the cause of the state change was contact between the ball and the cube. This effect was canceled in experiment 4, when an inanimate sphere replaced the intentional agent. Altogether, results suggest that, in the infants’ eyes, a ball (an inanimate object) has the power to cause an arbitrary state change, but only if it inherits this power from an intentional agent. Eight-month-olds are thus capable of representing complex event structures, involving an intentional agent causing a change with a tool.

---

(Post Starts Here)

Overview

The study looks at how babies understand tools and cause-and-effect relationships. It starts by checking if babies can tell when one object makes another move by touching it. This was done by showing 8-month-old babies videos where a ball moves another ball after hiding behind a screen. Surprising situations, where what the babies expect doesn't happen, make their pupils bigger. This shows they're surprised, which helps researchers understand if the babies know about cause and effect.

The research then looks at whether babies think a tool (like a ball) can make something else (like a cube) change, like starting to flicker because someone used the tool on purpose. Earlier studies thought babies couldn't understand such ideas with just inanimate objects. But this study suggests babies might recognize that when a person (or something acting like a person) uses a tool, it can cause changes.

In a series of experiments, the researchers showed babies different scenarios with balls and cubes to see how they reacted to the interactions. They found that babies seemed surprised when the expected contact between objects didn't happen, which suggests they were thinking about the connections between the objects' movements and the outcomes.

Empirical

The study looked at how babies react to different types of events using their eye reactions, specifically how their pupils change size. Researchers did four experiments to see how 8-month-old infants respond to two main kinds of events: when one object touches another (Contact events) and when it stops short of touching (Gap events). They were particularly interested in seeing if babies could understand hidden parts of these events and how they reacted differently to them.

Experiment 1

• Babies watched a simple show with two balls.
• Sometimes, the first ball hits the second one (Contact). Other times, it stopped just before reaching the second ball (Gap). Most times (60%), something blocked the view so the babies couldn't see what happened to the balls (Occluded) to see if babies could guess what happened when they couldn't see the balls (using their imagination).
• Finding: Babies showed more surprise (their pupils got bigger) when the balls didn't touch (Gap) compared to when they did touch (Contact). This means they were more interested or found it more unexpected when the balls didn't touch.

Experiment 2 (Control)

• The same as Experiment 1, but this time, babies only saw the Contact and Gap events without any Occluded events to check if babies react differently when they always see what happens.
• Finding: This time, there was no difference in how surprised the babies were between when the balls touched or didn't touch. This suggests that seeing the events directly made both outcomes equally expected of them.

Experiment 3

• A new set of babies watched a different show involving a ball, a cube, and an "intentional agent" (something that seemed to move on its own).
• The ball either hit the cube (Contact) or stopped before reaching it (Gap), with some events being hidden (Occluded) like before to see if the babies could understand the intention behind the moving objects and how they would react to unexpected outcomes.
• Finding: Babies were more surprised (bigger pupil dilation) when the ball didn't hit the cube (Gap) than when it did (Contact), indicating they found the Gap events more unexpected.

Experiment 4

• Similar to Experiment 3, but the "intentional agent" was replaced with an inanimate sphere, removing the aspect of intention or purpose to test if the presence of an "intentional agent" affects how babies perceive and react to the events.
• Finding: This time, there was no significant difference in babies' reactions to the Contact or Gap events, suggesting that the "intentional agent" plays a role in how they interpret these interactions. (Notice how this result is similar to experiment 2, suggesting that the "intentional agent" plays a similar role to the "occlusion" effect).

Ultimatum

Key Findings:

• Infants distinguish between things that move on their own (agents) and things that don't (objects). Agents can cause changes directly or by using objects as tools.
• Tools are seen as a bridge between agents and objects. They can "inherit" the agent's ability to cause change.

Constraints on Causal Power:

• There are limits to what infants expect regarding how agents and tools work. They expect contact and immediate effect.
• Future studies might explore when infants understand tools that work without direct contact, like a remote control.

Early Understanding of Tools:

• By 8 months, infants already understand the concept of using tools for indirect actions. This is before they can perform these actions themselves.
• This suggests the concept of a tool and its use is a basic part of human cognition early in life.

Comparative Aspects:

• Humans are unique in creating complex tools. It's not clear if other animals, like monkeys or chimps, share the same understanding of tools, especially without clear causal connections.

0 comments

Comments sorted by top scores.