According to RFT, types of relational responding may also become generalized through multiple exemplar training. As mentioned previously, the degree to which this is true can only be determined by empirical research. Researchers in many different fields study factors and experiences that contribute to language acquisition and development. This research is difficult to conduct, however, and we have few complete answers.
Nevertheless, it does seem that many aspects of early language training resemble multiple exemplar training for relational responding. Of course, in the natural world, this training typically isn’t as clear, formal, or systematic as it is presented here. But it does seem to happen. This kind of training could be what allows humans to make relational responses based on contextual cues to do so (rather than the physical properties of the stimuli).