Using other minds as a window onto the world: guessing what happened from clues in behaviour.

Pillai et al. (2014) showed adults short silent video clips. Each clip showed a person reacting to something that had just happened off-camera.

Viewers had to pick which of four stories caused the reaction. Half the viewers had autism, half did not. The task tested "retrodiction"—working backward from a behavior to the hidden event.

The autism group picked the right story far less often than the comparison group. They could see the facial expressions and body language, but could not figure out what event had triggered them.

Kaland et al. (2007) already showed that youth with Asperger syndrome need extra time to read minds. Dhanya’s group shows the problem is not just speed; the backward step itself is shaky.

Loukusa et al. (2007) found that kids with ASD skip context when they explain language. Dhanya extends this to real-life social scenes: they also skip context when they explain reactions.

Huang et al. (2017) looked forward, not backward. Their preschoolers with ASD ignored a grown-up’s earlier intention when deciding what to copy. Together the two papers form a timeline: the same inferential gap shows up at four years old and still lingers in adulthood.

If a client melts down after recess, don’t assume they can tell you what went wrong on the playground. They may need you to replay the scene, label each cue, and fill in the missing cause. Build lessons that run events in reverse: show the reaction first, then reveal the trigger, then practice matching them. Teaching "backward" social reasoning may be as important as teaching prediction.