Over the past 50 years, scientists have made amazing discoveries about the origins of human language acquisition. Central to this field of study is the process by which infants' perceptual sensitivities gradually align with native language structure, known as "perceptual narrowing." Perceptual narrowing offers a theoretical account of how infants draw on environmental experience to induce underlying linguistic structure, providing an important pathway to word learning. Researchers have advanced perceptual narrowing theory as a universal developmental theory that applies broadly across language learners. In this article, we examine diversity and representation of empirical evidence for perceptual narrowing of speech in infancy. As demonstrated, cumulative evidence draws from limited types of learners, languages, and locations, so current accounts of perceptual narrowing must be viewed in terms of sampling patterns. We suggest actions to diversify and broaden empirical investigations of perceptual narrowing to address core issues of validity, replicability, and generalizability.