Amanda Brandone and Henry Wellman, who made the finding, used a methodological approach that regular readers of the Digest will be familiar with. This is the preferential looking time procedure, which exploits the fact that babies tend to look longer at something novel that grabs their interest.
One hundred and thirty-four babies in three age groups – eight, ten and twelve-months – were habituated to one of two versions of a video showing someone reaching for a ball. To say the babies were habituated to the video means they were shown it enough times that they grew bored.
One version showed a person reaching, with an arc-like movement, over a mini wall to pick up a ball. The key thing about this video was that it showed someone intending to make a direct reach for the ball. The other version showed the same movement but the person failed to quite reach the ball – so the intent was the same, but they had failed.
Next the babies watched two further alternating videos: both were similar to the first they’d seen, but this time the wall wasn’t there. In one, a person is seen reaching directly for the ball, with a straight, horizontal arm movement. In other words, his intent was to make a direct reach for the ball, just as in the earlier video. In the other, the person makes an arc-like reaching movement (similar to that seen earlier), even though no wall is in the way. So this person intended to make an indirect reach.
The key question was – which of the later videos would most grab the babies’ interest: the first one, which matched the intent in the earlier video (a direct reach for the ball), but was perceptually different, or the second video which was perceptually similar because of the arc-like movement, but which reflected a different intent (i.e. an indirect reach for the ball)?
The answer depended on which version of the first video the babies had seen.
When the first video showed someone successfully picking up the ball, all age groups subsequently spent more time looking at the later video showing an arc-like, indirect ball reach. This suggests that all the babies, from eight months and upwards, understood the intent behind the successful reach of the ball, and therefore found the later video showing an indirect reach far more interesting. (Yes, maybe they should get out more, but remember they’re only little).
By contrast, among the babies who saw the initial video version with an unsuccessful ball reach, only the 10- and 12-month-olds subsequently spent longer watching the later video showing the indirect, arc-like ball grab. This suggests that only the older babies understood that the person in the first video was to trying to directly reach the ball, even though he’d failed.
Taken altogether this research suggests that the ability of babies to understand the intent behind failed actions builds on their earlier ability to understand the intent behind successful actions.
In the researchers’ words: “…these data illustrate the early emergence of an intentional framework in at least one key instance of human action. Moreover, they show that this early intentional understanding of action appears later than, and potentially builds upon, a prior action- and object-based understanding.”
Amanda C. Brandone, Henry M. Wellman (2009). You Can’t Always Get What You Want: Infants Understand Failed Goal-Directed Actions. Psychological Science, 20 (1), 85-91 DOI: 10.1111/j.1467-9280.2008.02246.x