Echoes on the motor network: how internal motor control structures afford sensory experience

The multi-dimensional nature of vocal learning

How learning affects vocalizations is a key question in the study of animal communication and human language. Parallel efforts in birds and humans have taught us much about how vocal learning works on a behavioural and neurobiological level. Subsequent efforts have revealed a variety of cases among mammals in which experience also has a major influence on vocal repertoires. Janik and Slater (Anim. Behav.60, 1-11. (doi:10.1006/anbe.2000.1410)) introduced the distinction between vocal usage and production learning, providing a general framework to categorize how different types of learning influence vocalizations. This idea was built on by Petkov and Jarvis (Front. Evol. Neurosci.4, 12. (doi:10.3389/fnevo.2012.00012)) to emphasize a more continuous distribution between limited and more complex vocal production learners. Yet, with more studies providing empirical data, the limits of the initial frameworks become apparent. We build on these frameworks to refine the categorization of vocal learning in light of advances made since their publication and widespread agreement that vocal learning is not a binary trait. We propose a novel classification system, based on the definitions by Janik and Slater, that deconstructs vocal learning into key dimensions to aid in understanding the mechanisms involved in this complex behaviour. We consider how vocalizations can change without learning, and a usage learning framework that considers context specificity and timing. We identify dimensions of vocal production learning, including the copying of auditory models (convergence/divergence on model sounds, accuracy of copying), the degree of change (type and breadth of learning) and timing (when learning takes place, the length of time it takes and how long it is retained). We consider grey areas of classification and current mechanistic understanding of these behaviours. Our framework identifies research needs and will help to inform neurobiological and evolutionary studies endeavouring to uncover the multi-dimensional nature of vocal learning. This article is part of the theme issue 'Vocal learning in animals and humans'.

Sensorimotor Integration Can Enhance Auditory Perception

Whenever we move, speak, or play musical instruments, our actions generate auditory sensory input. The sensory consequences of our actions are thought to be predicted via sensorimotor integration, which involves anatomical and functional links between auditory and motor brain regions. The physiological connections are relatively well established, but less is known about how sensorimotor integration affects auditory perception. The sensory attenuation hypothesis suggests that the perceived loudness of self-generated sounds is attenuated to help distinguish self-generated sounds from ambient sounds. Sensory attenuation would work for louder ambient sounds, but could lead to less accurate perception if the ambient sounds were quieter. We hypothesize that a key function of sensorimotor integration is the facilitated processing of self-generated sounds, leading to more accurate perception under most conditions. The sensory attenuation hypothesis predicts better performance for higher but not lower intensity comparisons, whereas sensory facilitation predicts improved perception regardless of comparison sound intensity. A series of experiments tested these hypotheses, with results supporting the enhancement hypothesis. Overall, people were more accurate at comparing the loudness of two sounds when making one of the sounds themselves. We propose that the brain selectively modulates the perception of self-generated sounds to enhance representations of action consequences.

Learning to Expect: Predicting Sounds During Movement Is Related to Sensorimotor Association During Listening

Sensory experiences, such as sound, often result from our motor actions. Over time, repeated sound-producing performance can generate sensorimotor associations. However, it is not clear how sensory and motor information are associated. Here, we explore if sensory prediction is associated with the formation of sensorimotor associations during a learning task. We recorded event-related potentials (ERPs) while participants produced index and little finger-swipes on a bespoke device, generating novel sounds. ERPs were also obtained as participants heard those sounds played back. Peak suppression was compared to assess sensory prediction. Additionally, transcranial magnetic stimulation (TMS) was used during listening to generate finger-motor evoked potentials (MEPs). MEPs were recorded before and after training upon hearing these sounds, and then compared to reveal sensorimotor associations. Finally, we explored the relationship between these components. Results demonstrated that an increased positive-going peak (e.g., P2) and a suppressed negative-going peak (e.g., N2) were recorded during action, revealing some sensory prediction outcomes (P2: p = 0.050, ηp2 = 0.208; N2: p = 0.001, ηp2 = 0.474). Increased MEPs were also observed upon hearing congruent sounds compared with incongruent sounds (i.e., associated to a finger), demonstrating precise sensorimotor associations that were not present before learning (Index finger: p < 0.001, ηp2 = 0.614; Little finger: p < 0.001, ηp2 = 0.529). Consistent with our broad hypotheses, a negative association between the MEPs in one finger during listening and ERPs during performance of the other was observed (Index finger MEPs and Fz N1 action ERPs; r = −0.655, p = 0.003). Overall, data suggest that predictive mechanisms are associated with the fine-tuning of sensorimotor associations.

Building a state space for song learning

The songbird system has shed light on how the brain produces precisely timed behavioral sequences, and how the brain implements reinforcement learning (RL). RL is a powerful strategy for learning what action to produce in each state, but requires a unique representation of the states involved in the task. Songbird RL circuitry is thought to operate using a representation of each moment within song syllables, consistent with the sparse sequential bursting of neurons in premotor cortical nucleus HVC. However, such sparse sequences are not present in very young birds, which sing highly variable syllables of random lengths. Here, we review and expand upon a model for how the songbird brain could construct latent sequences to support RL, in light of new data elucidating connections between HVC and auditory cortical areas. We hypothesize that learning occurs via four distinct plasticity processes: 1) formation of 'tutor memory' sequences in auditory areas; 2) formation of appropriately-timed latent HVC sequences, seeded by inputs from auditory areas spontaneously replaying the tutor song; 3) strengthening, during spontaneous replay, of connections from HVC to auditory neurons of corresponding timing in the 'tutor memory' sequence, aligning auditory and motor representations for subsequent song evaluation; and 4) strengthening of connections from premotor neurons to motor output neurons that produce the desired sounds, via well-described song RL circuitry.