Abstract Researchers have examined differences between free‐ and restricted‐operant teaching arrangements while equating response and reinforcement rates. Preliminary data suggest that nonhuman organisms acquire novel discriminations more quickly under free‐operant arrangements (Hachiya & Ito, 1991). In contrast, Bulla et al. (2024) found that humans learned novel discriminations more efficiently under restricted‐operant arrangements. This study systematically replicates Bulla et al. (2024). We taught participants to say the corresponding numeral 0–10 for Hindi, Eastern Arabic, and Mandarin numbers. We assigned each number set to a free‐operant, restricted‐operant, or control teaching arrangement. The study assessed the effects of each arrangement across two phases: (1) acquisition and (2) frequency building. Additionally, data were collected on measures of fluency outcomes. Results suggest that participants acquired novel discriminations more quickly under restricted‐operant arrangements. No major differences emerged in fluency outcomes when response–reinforcer relations remained equal. Procedural modifications are discussed to clarify distinctions between basic and translational findings.