Theory mind, or mind reading, refers participants’ ability to infer others’ mental states. Mental states cannot be directly observed, so they have to be indirectly inferred. Theory of mind, especially the performance as being measured by the “false-belief task” (REF), is traditionally regarded as an advanced ability that is unique to human older than 4 years old (REF). Non-human primates and human children younger than 3 years lack such ability. New experimental techniques, however, tell a different story. Different theories have been put forward to explain this discrepancy. This paper intends to adjudicate between these theories with the paradigm pioneered by Kovács et al. (2010).
A typical false-belief task involves two agents such as Anne and Lilly, two containers such as a box and a basket, and an object such as a marble. First, when Lilly was still in the room Anne put the marble into the box. Second, after Anne left the room Lilly moved the marble from the box into the basket. Finally, Anne returned to the room and participants were asked to determine where Anne would go to look for her marble. Given this experimental setting, Anne believed that the marble was in the box, but the marble was actually (or more precisely be believed by the participants) in the basket. Anne’s belief about the location of the marble was different from the actual location of the mable. So Anne’s has a false-belief about the location of the marble. If participants understand the diversity between different beliefs and know that an agent’s behavior is guided by the agent’s own belief, rather than by others’ beliefs or by the reality, they shall know that Anne would go to the box, rather than the basket to look for the marble.
If participants’ answer was the box, as being observed in human adults (REF), it means that they has the ability to distinguish between their own belief and the reality, i.e., they have the mind reading ability. If participants’s answer was the basket, as being observed in children younger than 4 years-old (REF), however, the story becomes a bit more complex. Different reasons could be used to explain these seemly “wrong” responses. One possible reason is that children younger than 4 years old lack the mind reading ability. Another possible reason is that the so-called explicit false-belief task is insensitive or is too difficult for children to elicit the intended behavior, even though they have the ability to understand others’ mind.
One group of the evidence supporting the latter viewpoint originated from the seminal research conducted by Kovács et al. (2010). In their experimental design, each trial comprises a wall, an agent, and an object. First, the agent hid the object behind the wall. Second, in the presence of the agent, the object moved out from the behind of the wall and left the screen. Third, with the absence of the agent, the object re-entered the screen and hid back to the behind of the wall. Finally, in the presence of the agent, the wall was lowered down, showing that the object was still laid there. The dependent variable for infants was the looking time to the last frame of the video, and for adults was the response latency when they were asked to press a key as soon as possible once they see the object behind the wall.
The logic behind this paradigm is that if the observed outcome was different from the predicted outcome, then the violation of expectation (REF) would result in longer fixations and longer response latencies. The violation of expectation was induced by the conflict of tbe beliefs between the agent and the participants. Given the experimental manipulation, the agent believed that the object was not behind the wall, but the participants believed that the object was behind the wall. The actual result of the video confirmed the belief of the participants but negated the belief of the agent.
One point we should highlight was that the task does not explicitly require participants to consider the belief of the agent. If participants were ignorant of the agent’s belief, then their responses would be the same regardless of the agent’s belief. If participants had implicitly and automatically represented agent’s belief, however, they would realized the conflict of their own belief and the agent’s belief. This violation of the belief status would confuse the participants and result in longer reaction times. Henceforth, the longer fixation and longer response latency would imply that the belief of the agent was automatically considered even if they were not required to do so. That was the result being observed by Kovács et al. (2010).
This is not the end of the story, however. This general conclusion that children younger than 4 years’ old and non-human primates lack mental reading ability have recently challenged by several newly developed experimental paradigms.