Proprioceptive signals from tongue positions are not strong, meaning it can be challenging to teach the articulation of speech sounds to language learners. Some of the most prevalent speech sound disorders in American English involve the rhotic /ɹ/. Speech-language pathologists provide modeling and feedback to their clients during treatment. Since tongue shape is not readily visible during rhotic production, clinicians rely heavily on verbal descriptions and auditory feedback to explain sounds to their clients. It is currently unclear how closely these approaches mirror or depend on veridical descriptions of tongue shape. An example from musical education provides an instructive backdrop: Although clarinet teachers frequently describe raising the tongue body for high notes, the tongue body must physically lower [Lulich et al., 2017 JASA]. Students successfully learn to play high notes regardless of the accuracy of the description, but would clarinet instruction be more effective if it reflected physical reality? Analogous issues and questions can be raised in the context of speech-language pathology. This study presents preliminary results of an investigation of student-clinicians’ self-awareness of tongue position during rhotic sounds.