Three experiments were implemented to test two potential sources of reinforcement for emergent Naming. In Experiment I, a time-lagged delayed multiple probe design was implemented to test the effects of multiple exemplar instruction across listener responses (match and point) with an echoic component across three participants diagnosed with developmental disabilities. The dependent variable consisted of the number of correct responses for untaught listener and speaker behavior following match-to-sample instruction. The independent variable was the rotation of match and point instruction with an echoic component. Increases in untaught listener and speaker behaviors as a function of independent variable were observed. Echoic behavior affected the joining of listener and speaker responses and the acquisition of Naming. The source of reinforcement for the joining of listener and speaker behavior as a function of echoic responding was unclear. Experiment II and III tested the effects of a second order classical conditioning procedure on the acquisition of Naming as a potential source of reinforcement for the joining of listener and speaker behavior and Naming. Participant A did not acquire full Naming in Experiment I, and was selected to participate in Experiment II. Three participants were selected for Experiment III. In both Experiment II and III, the correct numbers of untaught listener and speaker responses were recorded as the dependent variable across instructional and incidental Naming probes. Instructional Naming presented tacts for a set of stimuli during match-to-sample instruction. Incidental Naming consisted of the experimenter tacting the stimuli in the presence of the participants. The second order classical conditioning procedure, the independent variable, consisted of stimulus-stimulus pairing of conditioned stimuli (visual or vocal speech) with neutral stimuli (visual or vocal speech). A time-lagged multiple probe design was used to test the effects of the second order classical conditioning procedure. Significant increases in correct untaught listener and speaker responses across instructional and incidental Naming probe conditions. The second order classical conditioning procedure functioned as a source of reinforcement for the joining of listener and speaker behaviors.