In addition to representing a main source of data in linguistic research, example sentences are a core vehicle for linguists in teaching a wide range of phenomena to our students. However, the content of these sentences often reflects the biases of the researchers who construct them: referents are typically given Anglocentric proper names like John and Mary, reflecting (at least implicitly) dominant white culture and conformity to heteronormative gender roles. To support linguists in shifting these practices, we present the Diverse Names Database, a database of 78 names from a variety of languages and cultures, confirmed with native speakers. We outline the goals for the project, introduce our process of developing and adjusting the design, and present some additional issues and reflections for consideration, such as how to use the database as one component of an affirming, anti-racist, and gender-equitable linguistics pedagogy. We aim to generate meta-level discussions about disciplinary conventions and canons, and to challenge the idea that underlying linguistic structures are, or should be, the only things of relevance when constructing example sentences. How we teach linguistics is part of how we practise it, and how we do both matters to the composition and direction of the field.