 Academic research is a wonderful and important enterprise to be a part of. And the purpose of research of universities is to create and disseminate knowledge. And humanists and scientists use many kinds of sources, like texts and molecules, to understand society, history, people, and how the world works. And my own work is devoted to a field called developmental science. We study the behavior, brains, and learning capacities of young children to understand what human cognition is and how our experiences bring it to life. And at the Princeton Baby Lab, we study the intersection of two signatures of human cognition. On the one hand, our species prodigious attention, memory, perception, and pattern detection abilities. And on the other, our propensity for language and communication with others. The product of these signatures, efficient learning, and exchange of information, is one of the greatest demonstrations of the power of the human mind. And in particular, we study how babies break into the structure of the world. And I hope to show you why proficiency in finding structure is important. So what do I mean by this? I mean the process of finding regularities in patterned input, such as finding words in language. And learning patterns is critical. And generalizing what you know to new experiences is very important too. And it's a hallmark of language in particular, because language is filled with patterns and rules that govern how words are combined. Language learning seems impossibly complex, given that our eyes and our ears are bombarded with millions of bits of data at every moment. The eardrum, a thin, stretched membrane, just vibrates with fluctuations of air pressure, and we end up knowing fine-grained information about tens of thousands of words and how they go together. How do we understand infants' early stabilities to break into structure, right? Data come in, we detect structure, and we learn. Well, lab studies are a useful way to start. So in one interesting study, seven-month-old infants heard sequences of syllables that followed a particular pattern, like an AAB pattern. Le, le, di, go, go, la. And then the researchers measured how long infants wanted to listen to new syllables that either matched this pattern, like si, si, ru, or did not match this pattern, like si, ru, si. And infants listened longer to this novel ABA pattern. They were surprised, so to speak. They had learned something about this pattern. But critically, infants consistently failed to do this with non-speech sounds. If you do the identical study with sine wave tones or music notes, infants fail to learn. They don't show any element of surprise when they hear this violation of the pattern they learned. This led to the conclusion that speech has a privileged status in early learning, perhaps resulting from speech-specific adaptations that evolved with our capacity for language. But we tested a different hypothesis, that learning patterns is engaged by communication in general, not by speech, per se. And infants love people. They love their faces, their sounds, the fact that they move. So maybe their massive experience watching people use speech in dynamic ways is responsible for enabling the learning of patterns. So we actually tricked babies into thinking that tones could be used to communicate. We showed infants a short video of two women talking, one in English, one responding in sine wave tones, the other responding in English. And this brief video of people changed the landscape of their learning. They successfully learned the inherent structure in sequences of tones, just as well as they did with speech. This led us to conclude, based on various control and replication studies, that communicative signals grab our attention. And in doing so, catalyze a fundamental achievement of early learning, the detection of structure. And this is just one example of how infants find structure in the world. Our field has shown collectively that the moment-to-moment detection of structure and the processing of incoming information has cascading effects on later learning. The first 1,000 days of life have consequences, troubling consequences on outcomes, because learning is fragile and it's subject to the environment. So for example, in the United States, children from wealthier families here on average three times as much language as children from poor families. This gives some kids an enormous boost in discovering the structure of language and leaves other kids behind. And this brings me to another valuable contribution of our field. We are at a turning point in our ability to conduct rigorous interventions to understand how to foster successful learning. One example is the 30 million words project based at the University of Chicago. We are conducting a parent-directed intervention aimed at reducing disparities in language experience and reducing gaps in school readiness. We're using a randomized controlled trial design and we are providing an in-home course in either child development or nutrition with 200 families in poverty in Chicago. The goal is to see if we can enhance the quality of parent-child interactions and accelerate the early detection of structure. And labs all over the world, including here in China, are equipped with theories and methods to understand how human infants learn so well and also why some children are hindered in this pursuit, like children with autism or children raised in institutions who often do not have the same access to enriching experience. And developmental science allows basic and translational science to coexist productively. And key to this is that infants' learning trajectories diverge more and more over time. So if we're looking for the greatest return on investments in human capital, there's no time like infancy. If we can uncover how experience is structured and how infants discover structure in its complexities, we will be well positioned to unearth the consequential variable and still largely mysterious human ability to learn.