When can children work out a cause they cannot see?
By age 4, children used the sound a hidden ball made to find it, but only when the sound could plausibly have been caused by the ball, showing they reason about unseen causes rather than just learning associations.
Source
Inferring Unseen Causes: Developmental and Evolutionary Origins
Study at a glance
- Design
- Human experiment — Five experiments: between-subjects causal vs arbitrary event order with identical stimuli (Exp 1), a temporal-proximity control (Exp 2), chimpanzee versions (Exp 3-4), a simpler shaken-boxes task (Exp 4) and a causal-question version for 3-year-olds (Exp 5)
- N
- N=129 · Main Experiment 1: 129 children aged 3-6 (65 causal, 64 arbitrary); Exp 2: 40 children aged 4-5; Exp 4: 48 children aged 3-5 and 11 chimpanzees; Exp 5: 28 3-year-olds; Exp 3: 6 chimpanzees then 11 in a second phase
- Population
- Preschool children aged 3-6 in Scotland; chimpanzees at Leipzig Zoo for the comparative experiments
- Outcome
- Proportion of 20 (or 10) trials on which the hidden ball or reward was found, compared with chance and between conditions; children's verbal explanations
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Children watched a ball containing a sticker dropped into an opaque forked tube ending in two boxes, and heard either a metallic 'ding' or wooden 'clack'. In the causal condition the sound came after the ball fell, as if made by the ball landing; in the arbitrary condition exactly the same sound came before the drop, so it still predicted the box but could not have been caused by the ball. Follow-ups checked whether the sound's timing alone mattered, tested chimpanzees on the same and simpler tasks, used a simpler shaken-box task, and asked 3-year-olds a causally worded question ('Which box did I shake?').
What they found
4-year-olds found the ball above chance only in the causal condition and did significantly better there than in the arbitrary condition; 5-year-olds showed a similar non-significant trend, 6-year-olds succeeded in both conditions, and 3-year-olds failed both. With purely arbitrary recorded sounds, the order of sounds made no difference, ruling out a simple timing explanation. Children who explained that the boxes made different sounds performed much better (a proportion correct of 0.77 versus 0.52). Chimpanzees stayed at chance in every version, and 3-year-olds succeeded (0.67 correct) only when asked the causally framed question.
The limits
What it doesn't show
The failures of 3-year-olds and chimpanzees are null results that could reflect working-memory or attention demands of tracking brief sounds rather than an inability to reason causally, as the authors acknowledge. The chimpanzee samples were tiny (6 and 11 animals) and most animals developed side biases, so the comparative conclusions are weak. Age groups within Experiment 1 contained only about 16 children per condition, the 5-year-old difference was not significant, and even successful 4-year-olds were far from ceiling. It is also unclear whether the causal question in Experiment 5 scaffolded reasoning or simply turned the task into sound-to-box matching.
Key terms
- Causal reasoning
- Inferring what produced an effect based on a representation of how causes work, rather than only on learned co-occurrence.
- Associative learning
- Learning that two events tend to occur together, allowing prediction without any understanding of the mechanism linking them.
- Arbitrary condition
- A control in which the same cue predicts the outcome equally well but could not physically have caused it, here because the sound came before the ball dropped.
- Dual representation
- Treating something both as an object or event in itself and as a symbol standing for something else; proposed to explain 6-year-olds' success with arbitrary cues.
- Side bias
- Repeatedly choosing the same location regardless of cues, a common strategy when a subject cannot solve a two-choice task.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
In Experiment 1, what made the arbitrary condition 'arbitrary'?
Common questions
Why is the arbitrary condition such a strong control?
It uses exactly the same sounds, boxes and ball, and the sound predicts the location just as reliably; only the order of events changes. An associative learner should do equally well in both, so better performance in the causal condition points to causal reasoning.
Do these results prove chimpanzees cannot reason about hidden causes?
No. They failed, but negative results are hard to interpret; the authors suggest the need to remember transient sounds and integrate several pieces of knowledge may have overloaded attention and memory.
Why did 6-year-olds do well even in the arbitrary condition?
The authors think older children can treat an arbitrary cue as a symbol or signal, so any reliable predictor becomes usable even without a causal link.
More on Development