Search for: All records

Humans have an astonishing ability to extract hidden information from the movements of others. For example, even with limited kinematic information, humans can distinguish between biological and nonbiological motion, identify the age and gender of a human demonstrator, and recognize what action a human demonstrator is performing. It is unknown, however, whether they can also estimate hidden mechanical properties of another’s limbs simply by observing their motions. Strictly speaking, identifying an object’s mechanical properties, such as stiffness, requires contact. With only motion information, unambiguous measurements of stiffness are fundamentally impossible, since the same limb motion can be generated with an infinite number of stiffness values. However, we show that humans can readily estimate the stiffness of a simulated limb from its motion. In three experiments, we found that participants linearly increased their rating of arm stiffness as joint stiffness parameters in the arm controller increased. This was remarkable since there was no physical contact with the simulated limb. Moreover, participants had no explicit knowledge of how the simulated arm was controlled. To successfully map nontrivial changes in multijoint motion to changes in arm stiffness, participants likely drew on prior knowledge of human neuromotor control. Having an internal representation consistent with the behavior of the controller used to drive the simulated arm implies that this control policy competently captures key features of veridical biological control. Finding that humans can extract latent features of neuromotor control from kinematics also provides new insight into how humans interpret the motor actions of others. NEW & NOTEWORTHY Humans can visually perceive another’s overt motion, but it is unknown whether they can also perceive the hidden dynamic properties of another’s limbs from their motions. Here, we show that humans can correctly infer changes in limb stiffness from nontrivial changes in multijoint limb motion without force information or explicit knowledge of the underlying limb controller. Our findings suggest that humans presume others control motor behavior in such a way that limb stiffness influences motion.