In Stemmer (1992) it is shown that, with respect to practically all verbal items, correct speaker behavior with respect to a particular item requires the previous acquisition of correct listener behavior with respect to this item. Once the listener behavior has been acquired it can then be "transferred" into speaker behavior1. Therefore, the correct analysis of the processes that establish listener behavior is of primary importance for the analysis of verbal behavior, including speaker behavior. Stemmer (1992) contributes to this project by analyzing so-called ostensive learning processes. Very briefly, in these processes children hear expressions such as "bicycle," "giraffe," or "thunder" while "paying attention" to a particular stimulus. That is, the children are exposed to the pairing of a verbal stimulus (the ostensive expression) with a non-vocal stimulus. These pairing events establish (relatively) correct listener behavior with respect to the ostensive expressions. For example, if a child is exposed to the pairing of the verbal stimulus "This is a bicycle" with the presentation of a (salient) bicycle then, after the exposure to this pairing event, vocal stimuli such as "Give me a bicycle" will often evoke correct listener responses (e.g., Nelson & Bonvillian, 1973; see also Skinner's example of the listener behavior that is generated by an exposure to the ostensive pairing of the vocal stimulus of "Jones-plug" with a Jones-plug, 1957, p. 360). Ostensive learning processes are not operant conditioning processes. Rather, they are closely related to Pavlovian processes in the following sense: (a) a necessary, and often even a sufficient, condition for the learning effects to occur is the exposure to the pairing of two salient stimuli, one a verbal stimulus (e.g., the word "Jones-plug"), and the second, some other stimulus (e.g., the Jones-plug); and (b) the learning occurs without the systematic reinforcement of a specific response to the vocal stimulus.