PROJECT SUMMARY When Confucius said, “Tell me who are your friends, and I’ll tell you who you are,” he was noticing that how we behave and communicate is shaped by who we choose to hang out with every day. We constantly mimic the mannerisms and behaviors of friends and loved ones. Yet the neural basis of how we imitate, and more importantly who we choose to emulate and why, is largely unknown. Parrots provide a powerful yet untapped model system for social learning. Parrots, like humans and non-human primates, live in a specific type of ‘fission- fusion’ social network in which making and maintaining friendships is the key to fitness. Like humans, they selectively imitate and learn the names of their carefully selected companions. Here we aim to launch parrots as a new animal model in systems neuroscience. In aims 1, we will record neural activity in the vocal motor cortical output of the song system (nucleus AAC) in pairs of budgerigars engaged in courtship interactions. In these first- ever neural recordings form awake, behaving parrots, we are finding that AAC neurons exhibit premotor signals for vocalizations (as expected) and for expressive gestures such as silent kissing, head-bobbing and allogrooming. This joint vocal and gestural neural control, observed in human Broca’s area but not in songbirds – means that what was thought to be a songbird-like ‘song system’ is actually a more general system for social interaction. We next test the causal relationship between song system activity and social behavior. Inactivating AAC during courtship interactions will test if/how vocalizations and gestures degrade or lose their coordination (Aim 2.1). Inactivating frontal or posterior cortical inputs to AAC in bonded pairs will test the songbird-inspired idea that variability and order depend on distinct cortical pathways (Aim 2.2). For each inactivation experiment, a pair of interacting birds is conceptualized as a single dynamical system – and we will use machine learning guided behavioral analysis to quantify how vocalizations and gestures change (or do not) in both the inactivated and non-inactivated partner. Finally, in Aim 3 we will image dopamine (DA) release using fiber photometry and genetically encoded DA sensors. Pilot data demonstrate feasibility of DA imaging in singing birds. These experiments will test for the first time if DA signals, known to evaluate the quality of reward outcomes, similarly evaluate social outcomes. Courtship dynamics are perfect because gestural ‘requests’ to allogroom or ‘kiss’ are rejected or accepted with visually and acoustically obvious ‘consent’ or ‘deny’ signals. Males make hundreds of advances per day and use female feedback to learn – providing natural trial structure, within-session learning, and ‘events’ to which we can align simultaneously recorded male and female DA signals – which may or may not come into alignment as a pair ‘decides’ to bond or not. Budgerigar interactions resemble human conversations...