Understanding our social world is a complex task. Indeed, typical environments are noisy, visual information is often degraded, and sometimes sensory consequences are seemingly produced without a master. Nevertheless, our central nervous system (CNS), appears to have a resourceful way of helping us perceive our sensory words; it perhaps uses motor memories as a guide when explaining a sensory change.
At present, there is abundant research showing that sensory signals activate motor brain regions. For example, watching a peer produce simple hand actions has been shown to activate the brain regions (in the observer) that are involved in producing those movements. Hearing rehearsed piano notes can activate the fingers of both experts and recently trained non-musical controls. Even simply hearing speech appears to activate the muscles involved in producing those speech sounds, and this effect seems graded on the amount of work, for instance, the tongue must complete to generate those vocalizations.
This type of neurological response, where sensory signals activate motor pathways in the absence of overt movement, is often referred to as motor resonance (MR). Some early research suggested that the underlying MR brain mechanisms might be due to genetics, perhaps owing to our social and community-based evolution. Simply put, as social creatures, we have evolved to inherit a type of motor cell that can be activated by sensory information. However, other reports speculated that the response (and neural structure) could be generated by simple CNS learning systems, which encode the cells during a repeat performance. For instance, repeatedly locking an iPhone could associate that precise action with the sound it produces. As a result, the neurons controlling those actions become related to the sounds that are produced via activation. Therefore, when we hear those sounds in the environment, the corresponding motor cells are activated. Accordingly, these theories suggest Hebbian learning (or the idea cells that fire together wire together) is responsible for the integration at the neurological level of these sensory and motor information streams.
Another theory, which is based on the motor control research, addresses some shortcomings of the simple associative models, such as the underlying neurobiological complications of executing an action, and might indirectly explain MR. For example, processing sensory information takes time, and these sensorimotor delays are hard to reconcile via a simple association process. Additionally, sometimes it is difficult to even determine the origin of a sensory change, such as when trying to decide if your car or the one next to you is moving. Taken together, associating sensory and motor information is clearly more complicated than previously thought.
To overcome these issues, conventional motor control theories suggest the CNS generates an internal model of action by combining the sensory and motor information. While these are similar to the simple associations previously mentioned (e.g., a sensory-motor memory), the internal model mechanism uses a prediction component too. For instance, immediately prior to locking the iPhone, the internal model will determine what motor instructions are needed to achieve the desired outcome. Once those on are initiated, the system will calculate the impending sensory changes (e.g., what it should feel like to press the phone button, and then what noises should be produced by that action). Should there be slight errors in the initial movements, these predictions can help update the system (to accommodate the potential problem), without which might take too long to decide the appropriate correction. In any case, these predictions also help process the sensory information once it arrives.
At present, there are many reports of these type of predictive processes during action across species, including fish, birds, mice, and humans. As noted, there is data suggesting that sensory-motor memories are learned too. However, how these two components interact and generate a whole or working internal model is yet to be shown. Currently, there are investigations trying to determine if predictive processes affect how the sensory-motor memories are learned, and if so where they might be located in the brain.
Despite the suggestion that internal models can help sensory processing and action understanding, it should be mentioned that for unimpeded information the brain is likely to employ other heuristics that don’t require motor input. Nevertheless, these and similar motor control systems, while being an efficient way to execute a movement, might indeed help with our understanding of the social world given the noise and fuzziness we often experience.
These findings are described in the article entitled Echoes on the motor network: how internal motor control structures afford sensory experience, published in the journal Brain Structure Funct. This work was led by Jed Burgess from Deakin University.