The goal of this internship is describe children's input and outcomes based on extant transcriptions of the Tsimane', a forager-farmer society in the Bolivian Amazon. Quantitative studies have documented minimal child-directed speech (CDS) from Tsimane' adults to children and identified other children as substantial conversational partners to Tsimane' toddlers. However, qualitative dimensions of both linguistic inputs have largely remained unexplored. In particular, it remains uncertain whether "classical" CDS features, such as simpler and shorter utterances, are also employed by Tsimane' adults and other children when interacting with Tsimane' toddlers. Also, it is unclear whether Tsimane' children develop similarly to e.g. American children. Our team has collected naturalistic recordings from 24 participants (ages 6 to 68 months, median age 32 months) and transcribed sections of it. Together with Tsimane' informants, we also categorized vocalizations by speaker and addressee, distinguishing between the "key child" (the one wearing the recording device), adults, and other children. We will use Mean Length of Utterance (MLU) to investigate linguistic complexity in children's input and production.