The evolution of verbal behavior in children.
Complex language is one of the unique repertoires of the human species. Others include teaching and certain "types of imitation" (Premack, 2004), although these too may be pre-or co-requisites for certain functional uses of language. Over the last 40 years linguists have proposed theories and provided evidence related to their interpretation of the structure of language (Chomsky, 1959; Chomsky & Place, 2000, MacCorquodale, 1970; Pinker, 1999). Neuroscientists have identified neurological correlates associated with some aspects of language (Deacon, 1979, Holden, 2004). Behavior analysts have focused on the source of and controlling variables for the function of language as behavior per se (Catania, Mathews, & Shimoff, 1990; Greer & Ross, 2005; Michael, 1984; Skinner, 1957).
More recently, scholars have come to view human language as a product of evolution; "Linguists and neuroscientists armed with new types of data are moving beyond the non-evolutionary paradigm once suggested by Noam Chomsky and tackling the origins of speech head-on." (Culotta & Brooks-Hanson, 2004, p. 1315). The current work focuses on the evolution of both the non-oral motor and oral components of speech (Deacon, 1997; Holden, 2004), although some arguments are characterized necessarily more by theory than data.
Despite the evidence that primates and pigeons can be taught certa in features of verbal behavior (D. Premack & A. Premack; 2003; Savage-Rumbaugh, Rumbaugh, & Boysen, 1978; Epstein, Lanza, & Skinner, 1980), the speaker-as-own listener capability makes complex verbal behavior possible and may represent what is most unique about human verbal functions (Barnes-Holmes, Barnes-Holmes, & Cullinan, 2001; Lodhi & Greer, 1989; Horne & Lowe, 1996). Some suggest that oral communication evolved from clicking sounds to sounds of phonemes, and cite the extant clicking languages as evidence (Pennisi, 2004). It is likely that sign language and gesture predated both vocal forms; but it is the evolution of the spoken and auditory components of language that are seen as critical to the evolution of language. Some of these include changes in the anatomy of the jaw--Homo sapiens have more flexible jaw than did Neanderthals. Also, the location of the larynx relative to the trachea is different for Homo sapiens, and this anatomical feature made it possible for the humans to emit a wider range of speech sounds (Deacon, 1997). The combination of these anatomical changes, together with the identification of separate, but proximate, sites in the brain for speaking, listening, and imitation seem to be critical parts of what made spoken language possible (Deacon, 1997). The presence of these anatomical and physiological properties made it possible for the evolution of verbal functions through the process of cultural selection (Catania, 2001). The functional effects of speech sounds were acquired by the consequences provided within verbal communities. This latter focus is what constitutes the subject matter of verbal behavior.
The new foci on language, as an evolved anatomical and physiological capacity, do not necessarily suggest the existence of a universal grammar; nor, in fact, does it eliminate the possibility of an evolved universal grammar, as some argue (Pinker, 1999). Some of the linguistic neuropsychological searches for an evolved universal grammar now follow the PET and MRI trails and focus on identifying blood flow associated with the speech and hearing centers in the brain (Holden, 2004). Interesting and as important as this work may be, little, if any, is devoted to the function of language as behavior per se. Nor is it concerned with the biological or cultural evolution of verbal function in our species or in the lifespan of the individual, although anthropological linguists point to functions as the initial source. Only the research associated with Skinner's (1957) theory of verbal behavior as behavior per se, and expansions of the theory by contemporary behavior analysts, provide the means for analyzing how cultural selection gave rise to the function of language (Greer, 2002; Greer & Ross, in press; Hayes, Barnes-Holmes, & Roche, 2000; Lowe, Horne, Harris, & Randle, 2002). Currently, the linguistic, neuropsychological, and behavior analytic foci remain separate sciences, though they need not remain so (Catania, 1998). While the role of cultural selection in the evolution verbal behavior for the species remains theoretical, the development of verbal behavior within the ontogeny of the individual is empirically verifiable.
From Theory to Research
For decades after the publication of Skinner's (1957) book on verbal behavior, the majority of the publications on the theory remained theoretical. There is now a significant body of research supporting and expanding Skinner's theory of verbal behavior. We have identified over 100 experiments devoted to testing the theory and utility for educational purposes. There is an additional significant body of related work in relational frame theory that includes at least an equal number of studies (Hayes et al., 2000). In our program of research alone, we have completed at least 48 experiments (25 published papers, several in press, and recent dissertations) and a number of replications. Our particular research program was driven by our efforts to develop schools that provide all of the components of education based solely on teaching and schooling as a scientific endeavor. While the existing work in the entire corpus of behavior analysis provided a strong foundation for a science of schooling, much was still missing. Cognitive psychology offered a plethora of theories and findings, and when they were germane to our efforts, these findings proved to be operationally synonymous to those identified in behavior analysis. However Skinner's (1957) Verbal Behavior showed the way for a research program to fill in much of what was missing in the literature in a manner that allowed us to operationalize complex cognitive repertoires.
In our commitment to a thoroughgoing scientific approach to schooling, we needed functional curricula that identified repertoires of verbal operants or higher order operants, including "generative" or "productive" verbal behavior. Our efforts included using pre-existing conceptual and applied verbal behavior research, identifying the needs of children who were missing certain repertoires, and identifying the validity of untested components of Skinner's theory through new experiments done by others and us (Greer, McCorkle, & Williams, 1989; Selinske, Greer, & Lodhi, 1991). Through this process we have been able to meet real educational needs, or at least the most pressing needs--the recognition of which were missing in the existing science of behavior and cognitive psychology. Of course, these educational voids were also apparent in normative practices in education based on pre-scientific approaches that treat teaching as an art. We needed findings that worked in the day-to-day operation of our schools, if we were to educate the "whole child." Along the way, we discovered some interesting aspects of verbal behavior that may prove useful to a behavioral developmental psychology (Baer, 1970; Bijou & Baer, 1978; Gewirtz, Baer, Roth, 1958). Indeed, the evidence suggests that we have identified what Rosales-Ruiz and Baer (1996) described as "behavioral cusps"--in our case verbal behavior cusps. Rosales-Ruiz and Baer stated that,
A cusp is a change [a change in the capability of the child] that (1) is often difficult, tedious, subtle, or otherwise problematic to accomplish, yet (2) if not made, means little or no further development is possible in its realm (and perhaps in several realms); but (3) once it is made, a significant set of subsequent developments suddenly becomes easy or otherwise highly probable which (4) brings the developing organism into contact with other cusps crucial to further, more complex, or more refined development in a thereby steadily expanding, steadily more interactive realm. (Rosales-Ruiz & Baer, 1996, p. 166). [The italics in brackets were inserted into the quotation.]
Repertoires of Verbal Behavior for Instructional Purposes
First, applications of the research findings in verbal behavior in our CABAS[R] schools led to the categorization of children for instructional purposes according to levels of verbal behavior or verbal capabilities that we extrapolated from Skinner's analysis of the components of verbal behavior (Greer, 2002). (1) Traditional diagnoses or developmental constructs are useful for some inquiries, but they are not very useful for instructional purposes. The identification of the functional verbal capabilities of children, however, that we extrapolated from Skinner's work was very helpful. Skinner described the different verbal repertoires of the speaker and the relation of the speaker and listener in terms of his observations of highly literate individuals. These repertoires seemed to constitute what individuals needed to posses if they were to be verbally competent. Moreover, those verbal functions provided operational descriptions for most of the complex educational goals that had been prescribed by educational departments throughout the western world (Greer & Keohane, 2004; Greer & McCorkle, 2003). For educational purposes, the capabilities or cusps provided us with behavioral functions for a curriculum for listening, speaking, reading, writing, and the combinations that made up complex cognitive functions.
The verbal categorization proved useful in: (a) determining the ratio of instructors to students that would produce the best outcomes for students (Table 1), (b) identifying what existing tactics from the research worked for children with and without particular verbal capabilities (See Greer, 2002, Chapters 5 and 6), isolating the specific repertoires children could be taught given what each child initially brought to the table, and the development of a curricula composed of functional repertoires for complex human behavior. Most importantly, we identified the verbal "developmental cusps" (Rosales-Ruiz & Baer, 1996) or specific verbal capabilities we needed to induce, if we were to make real progress with our children. The categories provided a continuum of instructional sequences and developmental interventions that provided a functional approach to cognitive academic repertoires, and the recasting of state and international educational standards into functional repertoires of operants or higher order operants rather than structural categories alone (Greer, 1987, 2002; Greer & McCorkle, 2003). Each of the major verbal categories also identified levels of learner independence (i.e., operational definitions of autonomy) as well as what we argue are valid measures of socialization. Table 1 lists the broad verbal stages as we have related them to independence and social function.
Much of our work as teacher scientists is devoted to experimentally identifying prerequisites or co-requisites repertoires needed by each child to progress through the capabilities listed in Table 1. Once these were identified, we used or developed scientifically based tactics for moving children with the lack of a particular verbal capability from one level of verbal capability to the next level in the continuum.
When we found it necessary, and were able to teach the missing repertoires, the children made logarithmic increases in learning and emergent relations ensued. That is they acquired what has been characterized in the literature as behavioral cusps. As the evidence accumulated with individual children across numerous experiments, we also began to identify critical subcomponents of the verbal capabilities. As we identified more subcomponents, we worked our way inductively to the identification of the developmental components within the verbal capabilities suggested by Skinner. The quest led serendipitously to increased attention on the listener and speaker-as-own listener repertoires, a focus that began to be evident in the work of others also (Catania, Mathews, & Shimoff, 1990; Hayes, et al., 2000; Horne & Lowe, 1996). Table 2 lists the verbal capabilities and the components and prerequisites that we are beginning to identify as well as some of the related research.
It was evident that without the expertise to move children with language delays through a sequence of ever more sophisticated verbal capabilities or cusps, we could make only minimal progress. As we began to identify ways to provide missing capabilities, the children began to make substantial gains. As the magnitude of the differences became apparent in what the children were capable of learning following the attainment of missing repertoires, we came to consider the possibility that these verbal repertoires represented developmental verbal capabilities or verbal behavior cusps.
We have shown that certain environmental experiences evoked the capabilities for our children. However, we are mindful that providing particular prerequisite repertoires that are effective in evoking more sophisticated verbal capabilities in children with language disabilities or language delays does not necessarily demonstrate that the prerequisites are component stages in all children's verbal or cognitive development. While Gilic (2005) demonstrated that typically developing 2-year old children develop naming through the same experiences that produced changes in our children with verbal delays, others can argue effectively that typically developing children do not require specially arranged environmental events to evoke new verbal capabilities. A definitive rejoinder to this criticism awaits further research, as does the theory that incidental experiences are not required. See Pinker (1999) for the argument that such experiences are not necessary.
Milestones of the Development of Verbal Function: Fundamental Speaker and Listener Repertoires
Our rudimentary classifications of children's verbal development adhered to Skinner's (1957) focus on the verbal function of language as distinguished from a structural or linguistic focus. Skinner focused on antecedent and consequent effects of language for an individual as a means of identifying function, as distinguished from structure (Catania, 1998). Eventually, his theory led to a research program devoted to the experimental analyses of verbal behavior with humans. In a recent paper (Greer, & Ross, 2004) and a book in progress (Greer & Ross, in press), we have suggested that this research effort might be best described as verbal behavior analysis, often without distinction between its basic or applied focus. We have incorporated the listener role in our work, in addition to the speaker functions. While Skinner's self-avowed focus was the speaker, a careful reading of Verbal Behavior (Skinner, 1957/92, 1989) suggests much of his work necessarily incorporated the function of listening (e.g., the source of reinforcement for the listener, the speaker as listener). Our research on the role of the listener was necessitated by the problems encountered in teaching children and adolescents with language delays, of both native and environmental origin, to achieve increasingly complex cognitive repertoires of behavior. Without a listener repertoire many of our children could not truly enter the verbal community. We needed to provide the listener roles that were missing, but that were necessary if the repertoires of the speaker were to advance. Skinner made the point that a complete understanding of verbal behavior required the inclusion of the role of the listener (See the appendix to the reprint edition of Verbal Behavior, published by the B. F. Skinner Foundation, 1992, pp.461-470). Moreover, new research and theories based on Skinner's work have led to a more complete theory of verbal behavior that incorporates the role of the listener repertoire. These efforts include, but are not limited to:
* Research done by relational frame theorists (Barnes-Holmes, Barnes-Holmes, & Cullinan, 1999; Hayes, Barnes-Holmes, & Roche, B., 2000),
* Naming research by Horne and Lowe and their colleagues (Horne & Lowe, 1996; Lowe, Horne, Harris, & Randle 2002),
* Research on auditory matching and echoics (Chavez-Brown & Greer, 2004)
* Research on the development of naming (Greer, et al., 2005b)
* Research on conversational units and speaker-as-own-listener (Donley & Greer, 1993; Lodhi & Greer, 1989), and
* Research on learn units (Greer & McDonough, 1999).
Our levels of verbal capability incorporate the listener as part of our verbal behavior scheme (Skinner, 1989). The broad categories that we have identified to date are: (a) the pre listener stage (the child is dependent on visual cues, or, indeed, may not even be under the control of visual stimuli), (b) the listener stage (the child is verbally governed as in doing as others say) (c) the speaker stage (the child emits mands, tacts, autoclitics, intraverbal operants), (d1) the stage of rotating speaker-listener verbal episodes with others (the child emits conversational units and related components of learn units in interlocking operants between individuals), (d2) the speaker-as-own listener stage (the child engages in self talk, naming, speaker-as-own-listener editing function, and say-do correspondence), (e) reader (the child emits textual responding, textual responding as a listener and emergent joint stimulus control, and the child is verbally governed by text), (f) the writer stage (the child verbally governs the behavior of a reader for aesthetic and technical effects), (g) writer-as-own reader (the child reads and revises writing based on a target audience), and uses verbal mediation to solve problems (the child solves problems by performing operations form text or speech). Each of these has critical subcomponents and the subcomponents of the categories that we have identified to date are shown in Table 2.
The Listener Repertoire
In the verbal community a pre-listener is totally dependent on others for her care, nourishment, and very survival. Pre-listeners often learn to respond to a visual and tactile environment; but if they do not come under the control of the auditory properties of speech they remain pre-listeners. For example, in some situations they learn to sit when certain visual cues are present. It is often not the spoken stimuli such as "sit still, "look at me," or do this" to which they respond, but rather certain instructional sequences or unintentional visual cues given by teachers and caretakers. They do no respond to, or differentiate among, the auditory properties of speech as stimuli that evoke specific responses. When the basic listener repertoire is missing, children cannot progress beyond visual or other non-auditory stimulus control. However, substantial gains accrue when children achieve the listener capability, as we shall describe.
Auditory Matching. It is increasingly apparent, that children need to match word sounds with word sounds as a basic step in learning to discriminate between words, and even distinguish words from non-word sounds. While most infants acquire auditory matching with apparent ease, some children do not acquire this repertoire incidentally. Adults experience similar difficulties in echoing a new language.
Chavez-Brown & Greer (2003) and Chavez-Brown (2005) taught children who could not emit vocal verbal behavior or whose vocal speech was flawed to match pictures using BigMack[R] buttons as a pre-training procedure to teach them to use the apparatus. The teacher touched a single button set before her that had a picture on it and then touched each of the two buttons the students had in front of them (one with the target picture and one with a foil picture). Then students responded by depressing the button in front of them that matched the picture of the button in front of the teacher that had been touched by the teacher. Once the children mastered the visual matching task, used as a means to introduce them to the apparatus, we removed the pictures. In the next phase the children were taught to match the sound generated by the teacher's button (the buttons produced individual pre-recorded words or sounds). At this second stage, the depression of one of the students' buttons produced a sound and the depression of one of their buttons had no sound. Once they mastered matching sounds contrasted with no sound buttons, they learned to match words with non-word sounds as foils. Next, they learned to match particular words contrasted with different words. Finally, they learned generalized matching for words produced by pushing the buttons (i.e., they learned to match novel word sets with no errors). Our findings showed that children, who had never vocalized before, began to approximate or emit echoic responses under mand and tact-establishing operations when they mastered generalized word matching. Moreover, a second set of children, who had only approximations (i.e., faulty articulations), learned full echoics that graduated to independent mands and tacts. This matching repertoire may be an early and necessary step in the acquisition of speaking and may also be key to more advanced listening. See also correlations between auditory matching and the emissions of verbal operants identified by Marion et al. (2003) that suggested the auditory matching research we described above.
The Emersion of Basic Listener Literacy. When children have "auditory word matching" they can be taught the discriminative function needed to become verbally governed. Over the past few years, we found that children without listener repertoires reached a learning plateau and were no longer making progress in instruction beyond extensions of visual matching. We believe that children around the world who have these deficits are not making progress in early and intensive behavioral interventions. These children require inordinate numbers of instructional presentations, or learn units, and still do not make progress in acquiring repertoires that require verbal functions that are the very basic building blocks of learning. In an attempt to help these children become listeners, we developed an intervention that we call listener emersion (Greer, et al. 2005a). During listener emersion, we suspend all of the children's instructional programs and provide intensive instruction in responding to the discriminative acoustical properties of speech. This instruction continues until children's listener responses are fluent. (2)
In the listener emersion procedure, children learned to respond to words (i.e., vowel-consonant relations) spoken in person by a variety of individual voices as well as to voices recorded on tapes and other sources. By "fluent," we mean that the children learned to respond to four or more sets composed of five instructions such as "point to--," "match--", "do this," "stand up," and "turn around." The children also learned not to respond to nonsensical, impossible, or non-word vowel-consonant combinations that were inserted into the program as part of each set (i.e., "jump out the window," "blahblahblah"). These sets were presented in a counterbalanced format with criterion set at 100% accuracy. Next the children learned to complete the tasks at specified rates of accurate responding ranging from 12 to 30 per minute. Finally, they learned to respond to audio taped, mobile phone, or computer generated instructions across a variety of adult voices. Once the children's basic listener literacy emerged (i.e., the children met the listener emersion criteria), we compared the numbers of learn units required by each student to meet major instructional goals before and after listener emersion. The achievement of the objectives for the listener emersion procedure constitutes our empirical definition of basic listener literacy. This step insures that the student is controlled by vowel-consonant speech patterns of speakers. After acquiring basic listener literacy, the numbers of instructional trials or learn units the children required to achieve instructional objectives across the range of his or her instructional objectives decreased from four to ten times that which had been required prior to their obtaining basic listener literacy.
The Speaker Stage
Acquisition of Rudimentary Speaker Operants. In the late eighties, we identified procedures for inducing first instances of vocal speech that proved more effective than the operant shaping of spoken words as linguistic requests (Williams & Greer, 1989). That is, rather than teaching parts of words as vowel consonant blends, as had been the existing behavioral procedure (Lovaas, 1977), we arranged the basic establishing operations and obtained true mands and tacts using echoic -to-mand and echoic-to-tact procedures (Williams & Greer, 1989). Once true verbal operants were taught, the children used "spontaneous speech." The children came under the relevant establishing operations and antecedent stimuli (Michael, 1982, 1984, 1993) associated with mand and tact operants and related autoclitics, rather than verbal antecedent such as, "What do you want?" They did not require intraverbal prompts as a means of teaching pure tacts. In another procedure Sundberg, Loeb, Hale, and Eighenheer (2000/2001) evoked the emission of impure tacts and the emission of impure tacts and mands; these are necessary repertoires as well. In still other work Pistoljevic and Greer (2006) and Schauffler and Greer (2006) demonstrated that intensive tact instruction led to the emission of novel tacts and appropriate audience control.
Children who do not speak can be taught verbal behavior through the use of signs, pictures, or electronic speaking devices. Even so, we submit that speech is simply more useful; speech works in the community at large. When we are unable teach speech, we too, use these substitutes, although as our research has progressed there have been fewer children that we cannot teach to speak. The second choice for topography for us is electronic speaking devices as such devices supply the possibilities for speaker as own listener. The importance of speech becomes apparent when we reach the critical verbal repertoires of speaker-as own-listener and reader.
Although the use of the above procedures significantly increased the numbers of children we could teach vocal verbal operants, there were some children we still could not teach to speak. While we could teach these children to use substitute topographies for speech, the development of speech is critical for subsequent verbal capabilities. For those children who did not learn to speak using our basic echoic to-mand and echoic-to-tact procedures (Williams & Greer, 1989), others and we, designed and tested several tactics to induce first instances of speech. We taught children who had acquired fluent generalized imitation, but who could not speak, to perform chains of generalized imitation of large and small movement responses at a rate of approximately 30 correct per minute at 100% accuracy. These children were then deprived of preferred items for varying periods of time and were only able to obtain the items contingent on speech under conditions in which they first performed a rapid chain of generalized imitation (moving from large motor movements to fine motor movements related to touching their lips and tongue). As soon as the last motor movement step in the teaching chain was completed we offered the item under deprivation as we spoke its name. After several presentations as described, the children spoke their first echoic mands. Some of these children were as old as nine years of age and their first words were not separate phonemes but were mands like "baseball card," "Coke," or "popcorn." Once the echoic -to-mand was induced for a single word or words, other echoic responses were made possible and their independent mand repertoire was expanded--they acquired function. Follow-ups done years after these children spoke their first words showed that they maintained and expanded their mand and eventually their tact repertoires extensively (Ross & Greer, 2003). We currently think that the procedure acted to induce joint stimulus control across the two independent behaviors of imitating and echoing (see Skinner, 1957 for the important distinction between imitation and echoic responding). See-do joined a higher order class and a new behavioral cusp was acquired.
In a replication and an extension of this work, Tsiouri & Greer (2003) found that the same procedure could be used to develop tact repertoires, when the establishing operation was deprivation of generalized reinforcers. See Skinner (1957, page 229) for a source for the establishing operations for the tact. Moreover, tacts and mands could be evoked in tandem fashion when emission of the tact operants resulted in an opportunity to mand as a result of using the tandem procedures developed in Williams & Greer (1989) (Tsiouri & Greer, 2003).
The establishing operation is key to the development of these rudimentary operants (Michael, 1982, 1984, 1993). There appear to be three tested establishing operation tactics: (a) the interrupted chain (Sundberg, et al., 2001/2002), the incidental teaching procedure in which the incidental establishing operations opportunities are captured (Hart & Risely, 1975), and the momentary deprivation procedure (Williams & Greer, 1989). Schwartz (1994) compared the three procedures. She found them equally effective, although the momentary deprivation procedure resulted in slightly greater maintenance and required significantly less time. It is suggested that more powerful results may accrue if each of these establishing operations are taught in a multiple exemplar fashion providing the child with a range of establishing operations for controlling the emission of rudimentary operants. Still, other establishing operation tactic s are needed, such as identification of establishing operations for tacts described in Tsiouri and Greer (2003). Indeed, what is characterized in the literature as "naturalistic language" interventions, derived from Hart and Risely's incidental procedure, are essentially suggestions for capturing establishing operations as they occur in situ (McDuff, Krantz, McDuff, & McClannahan, 1988). The difficulty with relying solely on the capture of incidental establishing operations is that there are simply not enough opportunities to respond. There are now an abundance of tested tactics for evoking establishing operations in instructional sessions that can be used without waiting for an incidental occasion, although it is critical to capture incidental opportunities as well.
From Parroting to Verbal Operants. The stimulus-stimulus pairing procedure of Sundberg et al. (1996) evoked first instances of parroting of words as a source of automatic reinforcement. These investigators paired preferred events, such as tickling the children while the experimenters said words; the children began to parrot the words or sounds. Moreover, the children emitted the words in free play, suggesting that the saying of the words had acquired automatic reinforcement status. Yoon (1998) replicated the Sundberg et al. procedure, and after the parroting was present for her students, used the echoic-to-mand tactic described above (Williams & Greer, 1989), to evoke true echoics that, in turn, became independent mands. Until the parroted words were under the echoic to mand contingencies, the children were simply parroting as defined by Skinner (1957); however, obtaining the parroting as an automatic reinforcer made the development of true echoics possible. The emission of a parroting response may be a crucial first step in developing echoic responses and may be an early higher order verbal operant (3). The children in these studies moved from the listener to the speaker stage as a result of the implementation of extraordinary instructional procedures (See Sundberg & Partington, 1998 for an assessment and curriculum). Once a child has acquired a speaker repertoire the speaker-listener repertoire becomes possible. Speaker capabilities opened up extraordinary new possibilities for these children, as they did for our ancestors in the combined evolution of phylogenic capabilities in the context of capabilities evoked by cultural selection.
Transformation of Establishing Operations across Mand and Tact Functions. Initially, learning one form (e.g., word or words) in a mand or tact function does not result in usage of the form in the untaught function without direct instruction (Lamarre & Holland, 1985, Twyman, 1996). For example, a child may emit a word as a mand (e.g., "milk") under conditions of deprivation, such that the emission of "milk" results in the delivery of milk. But, the child cannot use the same form ("milk") under tact conditions (i.e., the emission of the word in the presence of the milk when the reinforcement is a social or other generalized reinforcement probability). The independence of these two functions has been reliably replicated in young typically and non-typically developing children; however, at some point most children can use forms acquired initially as mands and use the same forms as tacts, or vice versa. Some see this as evidence of something like a neurologically based universal grammar that makes such language phenomena possible (Pinker, 1999). Clearly, neural capacities must be present just as the acoustic nerve must be intact to hear. But, the unequivocal existence of a universal grammar does not necessarily follow; the source is at least as likely to lie in the contingencies of reinforcement and punishment and the capacity to be affected by these contingencie s in the formation of relational frames/higher order operants. One example of the acquisition of this verbal cusp or higher order operant is the acquisition of joint establishing operation control of a form in either mand or tact functions after learning only one function. When this verbal cusp is achieved, a child can use a form in an untaught function without direct instruction.
Nuzzolo-Gomez and Greer (2004) found that children who could not use a form learned in a mand function as a tact, or vice versa without direct instruction in the alternate function (Lamarre & Holland, 1985; Twyman, 1996a, 1996b), could be taught to do so when they were provided with relevant multiple exemplar experiences across establishing operations for a subset of forms. Greer, et al. (2003b) replicated these findings and we have used the procedure effectively with numerous children in CABAS schools. The new verbal capability doubled both incidental and direct instructional outcomes.
Speaker Immersion. Even after the children we taught had acquired a number of rudimentary speaker operants, some did not use them as frequently as we would have liked. Speaking had emerged; but it was not being used frequently, perhaps because the children had not received an adequate number of opportunities of incidental establishing operations. We designed a procedure for evoking increases in speaker behavior that we called speaker immersion (Ross, Nuzzolo, Stolfi, & Natarelli, 2006). In this procedure we immersed the children for whom the operants had already emerged in instruction devoted to the continuous use of establishing operations requiring speaking responses. All reinforcement was related to speaking and opportunities were provided throughout the day. As a result, the children's use of verbal operants dramatically increased as they learned to maximize gain with minimal effort. The children learned that it was easier and more efficient to get things done by speaking pure tacts and mands than by emitting responses that required the expenditure of more effort, thereby extending Carr and Durand's (1985) findings.
Milestones of Speaker and Listener Episodes: Interlocking Verbal Operants between Individuals
Verbal Episodes between Individuals
Verbal behavior is social as Skinner proclaimed, and perhaps one cannot be truly social without verbal behavior. A major developmental stage for children is the acquisition of the repertoire of exchanging speaker and listener roles with others--what Skinner (1957) called verbal episodes. A marker and a measure of one type of verbal episode is the conversational unit, while another type of verbal episode is a learn unit. We developed these measures as indices of interlocking verbal operants. No account of verbal behavior can be complete without the incorporation of interlocking verbal operants.
Epstein, et al. (1980) demonstrated verbal episodes between two pigeons. We argue that they demonstrated a particular kind of interlocking verbal operant that we identify as a learn unit. In that study, after extensive training, the researchers had two pigeons, Jack and Jill, respond as both speaker and listener in exchanges that simulated verbal episodes between individuals. Each pigeon responded as both speaker and listener and they exchanged roles under the relevant discriminative stimuli as well as under the conditions of reinforcement provided by each other's speaker and listener responses (a procedure also used in part by Savage-Rumbaugh, Rumbaugh, & Boysen, 1978). The pigeon that began the episode, the teacher pigeon, controlled the reinforcement in the same way that teachers deliver effective instruction (Greer & McDonough, 1999). That is, the teacher pigeon had to observe the responses of the student pigeon, judge its accuracy, and consequate the student pigeon's response. Premack (2004) argued that the lack of this kind of teaching observation in primates is evidence that this is one of the repertoires unique to humans. In the Epstein et al. study, special contingencies were arranged in adjacent operant chambers to evoke or simulate the teaching repertoire. Note that the pigeon that acted as a student did not emit the reciprocal observation that we argue needs to be present in the verbal episode we characterize as a conversational unit. In a conversational unit both parties must observe, judge, and consequate each other's verbal behavior.
We used the determination of verbal episodes as measures in studies by Becker (1989), Donley & Greer (1993), and Chu (1998) as well as related research by Lodhi and Greer (1989) and Schauffler and Greer (2006). The verbal episodes in these studies were measured in units and included a rotation of initiating episodes between individuals as well as a reciprocal observation accruing from reinforcement received as both a speaker and a listener. We called these episodes conversational units. A conversational unit begins when a speaker responds to the presence of a listener with a speaker operant that is then reinforced by the listener. This first piece of the verbal interaction is what Vargas (1982) identified as a sequelic. Next, the listener assumes a speaker role, under the control of the initial speaker who is now a listener. That is, the listener function results in the extension of sensory experiences from the speaker to the listener as evidenced by the speaker response from the individual who was the initial listener. The initial speaker then functions as a listener who must be reinforced in a listener function (i.e., the initial listener as speaker extends the sensory capacities of the initial speaker as a listener). A new unit begins when either party emits another speaker operant. Interestingly, in the cases of children with diagnoses like autism, we can now teach them a sequelic speaker function in fairly straightforward fashion using procedures described above. However, these children often have little interest in what the speaker has to say. The reinforcement function for listening is absent. We are currently working on procedures to address this problem.
Conversational units are essential markers and measures of social behavior and, we argue, their presence is a critical developmental milestone in the evolution of verbal behavior. By arranging natural establishing operations, Donley and Greer (1993) induced first instances of conversation between several severely delayed adolescents who had never before been known to emit conversation with their peers. Coming under the contingencies of reinforcement related to the exchange of roles of listener and speaker is the basic component of being social. Chu (1998) found that embedding mand operant training within a social skills package led to first instances of, and prolonged use of, conversational units between children with autism and their typically developing peers. Moreover, the use of conversational units resulted in the extinction of assaultive behavior between the siblings thereby extending Carr and Durand's (1985) finding.
Learn units are verbal episodes in which the teacher, or preprogrammed teaching device (Emurian, Hu, Wang & Durham, 2000), controls the onset of the interactions, the nature of the interactions, and most of the sources of reinforcement for the student. The teacher bases her responses on the behavior of the student by reinforcing correct responses or correcting incorrect response. The interactions provided in the Epstein et al. (1980) and the Savage-Rumbaugh et al. (1978) studies are learn units rather than conversational units as we described above. (See Greer, 2002, Chapter 2, for a thorough discussion of the learn unit and Greer & McDonough, 1999 for a review of the research).
Milestone of Speaker as Own Listener: Verbal Episodes "Within the Skin"
As Skinner pointed out, the speaker may function as her own listener as in the case of "self-talk." Lodhi and Greer empirically identified speaker as own listener in young typically developing children who engaged in self-talk while playing alone (Lodhi & Greer, 1989). This appears to be an early, if not the first, identification of conversational units in self-talk emitted by individuals under controlled experimental conditions. The developmental literature is replete with research on self-talk and its importance, but until the functional components defining self-talk were identified, self-talk remained essentially a topographical measure because the speaker and listener functions were not identified. It is very likely that speaker as own listener types of learn units are detectable also, although we have not formally tested for them except in our studies on print control that resulted in students acquiring self administration of learn units (Marsico, 1998).
We agree with Horne and Lowe (1996) that a speaker as own listener interchange occurs in the phenomenon that they identified as naming. Naming occurs when an individual hears a speaker emit a tact, and that listener experience allows the individual to emit the tact in a speaker function without direct instruction and further to respond as a listener without direct instruction. Horne and Lowe (1996) identified the phenomenon with typically developing children. Naming is a basic capability that allows children to acquire verbal functions by observation. It is a bi-directional speaker listener episode.
But what if the child does not have the repertoire? For example, matching, pointing to (both listener responses, although the point to is a pure listener response), tacting, and responding intraverbally to multiple controls for the same stimulus (the speaker response as an impure tact) are commonly independent at early instructional stages. This is the case because, although the stimulus is the same, the behaviors are very different. The child learns to point to red but does not tact (i.e., does not say "red" in the presence of red objects, or tacts and does not intraverbally respond to "What color?"). This, of course, is a phenomenon not understood well by linguists because they operate on the assumption that understanding is an automatic given--a human example of generative verbal behavior, if you will. It is a source of many problems in learning for typically developing and non-typically developing children, as well as college students who demonstrate differences in their responses to multiple -choice questions (selection responding) versus their responses to short answer or essay questions (production responding). At some point children can learn a match or point-to response and can emit a tact or intraverbal response without direct training. This is not, however, automatic for some children. Thus, we asked ourselves this question: If naming were not in a child's repertoire, could it be taught?
Induction of One Component of Naming. Greer, et al. (2005a) found that one could isolate experimentally a particular instructional history that led to naming for 2-dimensional stimuli (pictures) in children who did not initially have the repertoire. After demonstrating that the children did not have the repertoire for tacts, we provided a multiple exemplar instructional intervention with a subset of stimuli involving rotating match, point to, tact, and intraverbal responding to stimuli until the children could accurately do all of the responses related to the subset. We then returned to the initial set and a novel set as well and showed that the untaught speaker and listener repertoires had emerged.
These data suggested that the acquisition of naming, or one component of naming (i.e., going from listener to speaker) could be induced with multiple exemplar experiences. Naming is a generative verbal repertoire that Catania (1998) has called a "higher order class." The Relational Frame Theorists described this particular higher order operant as an incidence of transformation of stimulus function (Hayes, et al., 2000). Skinner referred to the phenomenon as responding in different media to the same stimulus (i.e., thematic grouping) and Relational Frame Theorists provided feasible environmental sources for this and related phenomena (i.e., multiple exemplar experiences). That is, a particular response to a single stimulus or category of stimuli when learned either as a listener repertoire or as a speaker repertoire is immediately available to the individual as a response without direct instruction once the individual has stimulus transformation across speaker and listener functions. We found that the naming repertoire emerged as a function of specific instructional experiences. This represents another case of the emergence of generative verbal behavior that is traceable to environmental circumstances. Fiorile and Greer (2006) replicated this finding. Naming also represents the acquisition of one of the speaker as own listener stages. When children have acquired it they have new verbal capabilities. Other types of generative behavior are traceable to multiple exemplar experiences, as we will discuss later.
Induction of Untaught Irregular and Regular Past Tense Responding. Still another case of speaker as own listener repertoires probably occurs in the emission of verb endings colloquially often associated with the cliche "kids say the darnedest things" (Pinker, 1992). We recently found that we could evoke untaught correct usage of regular and incorrect but "spontaneous" emission of irregular verbs (i.e., "he singed last night") as a result of multiple exemplar instruction with young children with developmental disabilities who could not emit either regular or irregular novel past tense forms without direct instruction (Greer & Yuan, 2004). The children learned to emit novel regular past tense forms without direct instruction and this abstraction was extended to irregular verbs. That is, they emitted incorrect irregular forms such as "he singed" as do young typically developing children. In a related study, Speckman (2005) found that multiple exemplar experiences also resulted in the emission of untaught suffixes as autoclitic frames for tacts. However, it is important to recognize that Pinker (1999) says the fact that children begin to use the correct irregular forms at some point and stop using the incorrect forms is a more important capability. He argues that there is no direct instruction leading to this revision in verb usage by typically developing children. But just as the initial incorrect usage has been traced to a sufficient set of experiences, it is possible that there are incidental sources of experience that make this change possible. We suspect that multiple experiences could induce this capability too, although further research remains to be done.
Milestones of Reading, Writing, Self-Editing: Extensions of the Speaker and Listener Repertoires
Reading involves textually responding (seeing a printed word and saying the word), matching various responses to the text as comprehension (printed stimulus to picture or actions, the spoken sound and all of the permutations and combinations of this relationship) (Sidman, 1994). At first glance, the reader stage appears to be simply an extension of the listener repertoire; however, on closer scrutiny, reading is necessarily an advanced speaker-as-own-listener repertoire because the reader must listen to what is read. Reading consists of speaker-listener relationships under the control of print stimuli, actions or pictures. Textually responding requires effortless rates of responding to print stimuli in order to "hear" the spoken word. After all, it was only after the Middle Ages that we began to read silently, and many religious and other ancient cultural practices still adhere to ceremonies in which one person reads aloud to an audience while the audience views the text.
The capacity to hear what one reads is important because the acoustical physical properties of sound allow more "bits" to be transmitted by sound than is possible with signs. For example, children who are deaf from birth have extreme difficulty developing reading comprehension beyond Grade 6 (Karchmer & Mitchell, 2003). There are special auditory properties of speech that allow a great deal of information or bits to be used for the benefits of the reader (aesthetic or functional), or at least this was the case before computers. Good phonetic instruction results in children textually emitting untaught combinations of morphemes and if those words are in their listener repertoire they can comprehend (See Becker, 1992 for the relevant research on multiple exemplar instruction and the emission of abstracted textual responses to untaught morphemes). However, even if a child can respond textually and thereby emit an accurate response to printed stimuli, and she does not have listener comprehension, the child "will not understand" what she has read (i.e., the child will be unable to match the sounds to a picture or action). We can textually respond to foreign language print aloud and have no idea about what we are saying. Thus, the listener component is key. For example, adolescents with multiple year delays in their reading achievement may not comprehend because they can not emit a textual response to a particular word or group of words, but once they hear a spoken version they immediately comprehend, because their listener vocabulary exceeds their textual repertoire. The listener component of reading is as important as the textual speaking component. Thus, a reader must be a reader-as-own listener, so to speak.
There is still a more basic component of reading that we identify as conditioned reinforcement for observing print and pictures in books. Tsai & Greer (2006) found that when they conditioned books such that 2 and 3 year old children chose to look at books in free time, with toys as alternate choices, the children required significantly fewer learn units to acquire textual responses. The book stimuli selected out the children's observing responses, and once the children were observing they were already closer to acquiring print stimuli as discriminative stimuli for textual responses. Thus, an early predictor for children's success in textually responding appears to be the conditioned reinforcement for observing book stimuli. Conditioned reinforcement for books may constitute a new capability. We currently also believe that pre-listener children who do not orient toward speakers and who are having listening and speaking difficulties may need to have unfamiliar and familiar adult voices acquire conditioned stimulus control for observing (Decasper & Spence, 1987). This too may be a crucial stage in the acquisition of listener repertoires.
Writing is a separate behavior from reading, and like the repertoire of speaking, represents a movement up the verbal scale. But writing from a functional verbal perspective requires that the writer affect the behavior of the reader; that is they must observe the effects of their writing and in turn modify their writing until the writing affects the behavior of the reader. In the case of technical writing the writer must provide technical information that affects the reader's behavior, ranging from influencing a shopper through the provision of a shopping list, to the provision of an algorithm that affects complex scientific decisions. Writing, as in the case of speaking, needs to be under the control of the relevant establishing operations if the writing is to be truly verbal. In several experiments we provided establishing operations for writing for students whose writing did not affect the behavior of the reader, using a tactic we call writer immersion. In the writer immersion procedure, all communication is done in written form for extended periods throughout the day. Written responses are revised until the reader responds as the writer requires. This procedure resulted in functionally effective writing, measured in effects on the behavior of readers, and improvements in the structural components of writing for the writer (grammar, syntax, vocabulary, punctuation, spelling) (Greer, Gifaldi & Pereira, 2003a; Keohane, Greer, & Mariano-Lapidus, 2004; Jadlowski, 2000; Madho, 1997; Reilly-Lawson & Greer, 2006). The experience taught the students to write such that they read as the target readers would read. The editing experience appears to evoke writer as own reader outcomes of self-editing, not unlike speaker as own listener (Jadlowski, 2000). This repertoire then appears to be an advanced speaker as own listener stage--one that requires one to read what one writes from the perspective of the target audience whose behavior the writer seeks to influence. Thus, like the reader function, the writer function builds on the speaker as own listener. Some individuals have difficulties in writing and reading that are probably traceable to missing components of the speaker, listener, or speaker as own listener components.
Complex Verbally Governed and Verbally Governing Behavior
Technical Writing. Another key component of the complex cognitive repertoires of individuals involves reading or being verbally governed by print for technical outcomes. Marsico (1998) found that teaching students to follow scripts under conditions that allowed the investigators to observe the control of the print over the students' responses resulted in students "learning to learn" new concepts in math and more complex reading repertoires by acquiring verbally governed responding from print sources. This repertoire allowed the students to be verbally governed by print. As this repertoire becomes more sophisticated it leads to the more complex repertoire of solving complex problems from algorithms as in the case of the following of decision protocols. Keohane and Greer (2005) showed that teacher scientists could perform complex data decision steps using algorithms based on the verbal behavior of the science, and this new repertoire resulted in significant improvements in the outcomes of the teachers' students. Verbal rules guided measurable responses involving data analysis, complex strategic analyses, and tactical decisions that were implemented with the teachers' students.
Nuzzolo-Gomez (2002) found that teachers who received direct learn units on describing tactics, or observed other teachers receive learn units on accurately describing tactics, required significantly fewer learn units to teach their children to achieve instructional objectives. Observations showed that the teachers' instruction was reliably driven by the verbal descriptions of the tactics they learned by direct or indirect instruction. These studies are analyses of the verbal behavior of scientists and the verbal stimulus control involved in either scientific complex problem solving repertoires suggested by Skinner (1957) and demonstrated in Keohane & Greer (2005), or the control of verbal behavior about the science over teacher performance as identified in Nuzzolo-Gomez (2002). We argue that these studies investigated observable responses that are both verbal and nonverbal and that such responses are directly observed instances of thinking.
While neuroscientists could probably locate electrical activity in the brain associated with our putative thinking responses, it is only the behavior outside the skin that distinguishes the electrical activity as thinking as opposed to some other event that might be correlated with the activity. Verbal stimuli control the complex problem solving, not the electrical activity. The electrical activity, although interesting, may be necessary, and important, but is not thinking per se. One might argue that the electrical activity is light in a black box; although we see within "the black box" we do not see outside of the black box. This is an interesting reversal of the black box puzzle. If the electrical activity were to begin before the relevant contingencies in the environment were to be in place the problem in the environment would not be solved.
One of the key components in writing is the process of spelling. Spelling involves two different and initially independent responses: (1) saying the letters for a dictated word and (2) writing the letters. At some point we do emit an untaught response after learning a single one of these behaviors (See Skinner, 1957, 1992, p.99). How does a single stimulus (i.e., hearing the word) come to control these two very different behavioral topographies of writing and orally saying the letters? Recently we found that for children who initially could not perform the untaught function, providing multiple exempla r instruction for a subset of words across the two responses under a single audited vocal stimulus resulted in these students acquiring the repertoire with novel stimuli (Greer, et al. 2004c). Like transformation of establishing operations for mands and tacts, and transformation of stimulus functions across speaker and listener in naming, the transformation of writing and saying in the spelling repertoires is still another environmental source for generative verbal behavior as an overarching operant or a higher order operant (Catania, 1998; Hayes et al., 2000). These repertoires consist of learned arbitrary relations between listening, speaking, and writing. It is not too far-fetched to infer that typically developing children acquire this joint stimulus control across independent responses as higher order operants or relational frames through multiple exemplar experiences. Such multiple exemplar experiences involve the rotation of writing and saying opportunities may occur incidentally rather than as a result of the programmed experiences we provided our children. Once the child has transformation of stimulus control over written and spoken spelling, only a single response need be taught.
In related research, Gautreaux, Keohane, & Greer (2003) found that multiple exemplar instruction also resulted in transformation of selection and production topographies in geometry. That is, middle school children who could not go from multiple -choice responding to production responding prior to multiple exemplar instruction, did so after an instructional history was created by multiple exemplar instruction across a subset of selection and production experiences. This study highlighted the difficulties experienced by some older children that may be due to a lack of prior verbal instructional histories. The replacement of missing verbal capabilities may be the key to solving instructional difficulties experienced later in life by individuals as they encounter more complex subjects. When an individual has difficulty with aspects of reading and writing, it is possible that the remediation of the difficulty only truly occurs when the missing capability is put in place. In effect, they have a missing or inadequate verbal developmental cusp. Inducing that cusp may solve the learning problems.
Aesthetic Writing. In an earlier section we described writing repertoires that were of a technical nature. Aesthetic writing has a different function than technical writing (Skinner, 1957). Aesthetic writing seeks to affect the emotions of the reader. To date, little empirical work has been accomplished with the aesthetic writing repertoire. A critical, if not the most basic component of aesthetic writing, is the writer's use of metaphors as extended tacts. Meincke (2005) and Meincke, Keohane, Gifaldi, and Greer (2003) identified the emergence of novel metaphorical extensions resulting from multiple exemplar instruction. This effort points to the importance of isolating and experimentally analyzing experiential components of aesthetic writing and suggests the role of metaphorical comprehension in reading for aesthetic effects. This also suggests that rather than teaching the aesthetics of reading through literary analysis as an algorithm, a student should have the relevant metaphoric experiences and perhaps these may be pedagogically simulated. It is likely that these metaphoric experiences provide the basis for the aesthetic effects for the reader. In order for the exchange to occur the target audience for the writer must have the repertoires necessary to respond to the emotional effects. Of course, the analysis of aesthetic writing functions is probably more complex than the analysis of technical repertoires, but we believe empirical analyses like the one done by Meincke et al. are becoming increasingly feasible. If so, the aesthetic and functional writer and reader repertoire may be revealed as new stages of verbal behavior.
From Experimental Effects to a Theory of Verbal Development
We believe we have identified several verbal repertoires that are key in children's development of successively complex repertoires of verbal behavior. Providing several of these repertoires to children who did not have them allowed these students to advance in their cognitive, social, technical, and aesthetic capabilities. As a result of this work we were increasingly persuaded that these levels of verbal capabilities did, in fact, represent empirically identifiable developmental cusps.
For our children the capabilities that they acquired were not tied to tautological relationships associated with age (Baer, 1970; Bijou & Baer, 1978; Morris, 2002). Age may simply provide a coincidental relation between experiences that bring about verbal capabilities and the probabilities of increased opportunities for those experiences. Hart and Risely (1996) showed that impoverished children who had no native disabilities, but who had significantly fewer language experiences than their more better off peers, demonstrated significant delays by the time they reached kindergarten. When children with these deficits in experience with language continued in schools that did not or could not compensate for their sparse vocabulary, these children were diagnosed as developmentally disabled by grade 4 (Greenwood, Delquadri, & Hall, 1984). It is not too farfetched to suggest that absence of the kinds of experiences necessary to evoke the higher order verbal operants or cusps that we have identified may also be part of the reason for these delays. We suggest that the presence of incidental multiple exemplar experiences provide the wherewithal for most typically developing children to seamlessly acquire the verbal milestones we described, probably because they have both the environmental experiences and neural capabilities (Gilic, 2004). For children without native disabilities who lack multiple exemplar experiences (Hart & Risely, 1996), as well as children with native disabilities who lack the necessary verbal capabilities, intensive multiple exemplar instruction has induced missing repertoires (Nuzzolo-Gomez & Greer, 2004). Such experiences probably result in changes in behavior both within and outside of the skin. Indeed biological evidence suggests that, "DNA is both inherited and environmentally responsive" (Robinson, 2004, p. 397. Also see Dugatkin, 1996 for research on the influence of the environment on changes in genetically programmed behavior affected by environmental events). What may be an arbitrary isolation of behavior beneath and outside the skin may dissolve with increased research in the environmental effects on both types of behavior.
Our induction of these repertories in children, who did not have them prior to instruction, suggests it is not just age (time) but particular experiences (i.e., environmental contingencies including contingencies that evoke higher order operants) that make certain types of verbal development possible, at least for the children that we studied. Intensive instruction magnified or exaggerated these experiences and provided our children with the wherewithal (i.e., verbal developmental cusps) to achieve new verbal capabilities. We speculate also that the induction of these verbal capabilities in children who do not have them prior to special experiences creates changes in neural activity. Of course, a test of this is the real challenge facing developmental neuroscience (Pinker, 1999). A joint analysis using the science of verbal behavior combined with instrumentation of the neurosciences might prove very useful in assisting children. Incidentally, such an analysis might also act to enrich academic debate towards more useful outcomes.
Tables 1 and 2 showed the levels of verbal functions for the pre-listener through the early reader stages in summary form. We described the evidence that has proved useful in our efforts to induce and expand progressively sophisticated verbal functions. The capabilities that we addressed were originally identified based on the responses of individual children; specifically they were based on our empirical tests for the presence or absence of the repertoires for individual children. In our educational work, when a particular repertoire was missing, we applied the existing research based tactics to provide the child with the repertoire. When we encountered children for whom the existing tactics were not effective, we researched new tactics or investigated potential prerequisite repertoires and related experiences that appeared to be missing for the child. The searches for possible prerequisite repertoires led to the identification of several subcomponents which when taught by providing subcomponent repertoires led to the emergence of verbal capabilities that were not present prior to our having provided the prerequisite instructional experience.
Summary of Identified and Induced Verbal Capabilities
We continue to locate other prerequisites and believe that there are many others that remain to be identified. Examples of rudimentary verbal functions that have been identified in the research include: (a) the emergence of better acquisition rates across all instructional areas as a function of teaching basic listening (Greer et al., 2005a), (b) the induction of parroting (Sundberg, et al., 1996) and then echoics that led to independent mand and tact functions (Yoon, 1996), and relevant autoclitics, for children with no speech or other verbal functions (Ross & Greer, 2003; Tsiouri & Greer, 2003), (c) transformation of establishing operations across the mand and tact function for children for whom a form taught in one function could not be used in an untaught function prior to multiple exemplar instruction (Nuzzolo & Greer, 2004), (d) the identification of interlocking speaker as own listener operants in self-talk with typically developing children (Lodhi & Greer, 1989), (e) induction of conversational units with children who had no history of peer conversational units (Donley & Greer, 1993), (f) the induction of naming in children who did not have naming prior to multiple exemplar instructional experience (Fiorile, 2005; Fiorile & Greer, 2006; Greer, et al., 2005b), (g) the emission of untaught past tenses for regular and irregular verbs as a function of multiple exemplar instruction (Greer & Yuan, 2004), (h) the emission of untaught contractions, morphemes and suffix endings as a function of multiple exemplar experiences or having children tutor using multiple exemplar experiences (i.e., observational learning through multiple exemplars) (Greer, et al., 2004a; Speckman, 2004), (i) faster acquisition rates for textual responses as a function of conditioning books as preferred stimuli for observing (Longano & Greer, 2006; Tsai & Greer, 2003), (j) and the induction or expansion of echoic responding a function of the acquisition of generalized auditory matching (Chavez-Brown, 2004).
The more advanced writer, writer as own reader or self-editing milestones are key complex cognitive repertoires. Research in this area includes: (a) teaching more effective writer effects on readers and structural responses of writing as a function of establishing operations for writing (Madho, 1997. Greer & Gifaldi, 2003; Reilly-Lawson & Greer, 2006), (b) the induction of rule governed or verbally governed responding and its effects on the verbal stimulus control of algorithms (Keohane & Greer, 2005; Marsico, 1998; Nuzzolo-Gomez, 2002), (c) the role of multiple exemplar instruction on the emergence of metaphors (Meincke et al., 2003), (d) transformation of stimulus function across vocal and written responding (Greer, et al., 2004c), and (e) the acquisition of joint stimulus control across selection and production topographies (Gautreaux, et al., 2003). These more complex repertoires appear to build on the presence of speaker as own listener capabilities.
While we are not ready to declare emphatically that the capabilities that we have identified experimentally, or by extrapolation from experiments, have been definitively identified as verbal developmental stages, the evidence to date shows that they are useful for instructional functions. Furthermore, they suggest possible natural fractures in the development of verbal function (4). For typically developing children, these fractures may occur as a result of brief experiences with exemplars. For some typically developing 2-year old children that we have studied, simply having a few experiences with exemplars going from listener to speaker, followed by single exemplars going from speaker to listener resulted in bidirectional naming for 3-dimensional stimuli that they did not have prior to those experiences that were separate and juxtaposed (Gilic, 2004). While our children with language delays required the rapid rotation across listener and speaker exemplars to induce naming, typically developing children may need only the incidental rotation of speaker listener experiences with single stimuli. It would appear that now that these generative or productive verbal capabilities have been traced to experiences for the children we have studied, the claim by some (Pinker, 1999) that productive or generative verbal capabilities is not traceable to learning experience is no longer credible.
Some of the research we described is not yet published and our references include papers presented at conferences or unpublished dissertations not yet submitted for publication. Thus, these are early days in our work on some of the stages. But it is important to note also that we have been on a quest for the last 20 years to remediate learning problems based on verbal behavior deficits in children with and without disabilities. The quest has moved forward based on progressively more complex strategic analyses as we stumbled on what we now believe may be developmental milestones in verbal behavior. We have replicated most of the effects we have identified with numerous children in our CABAS schools in the USA, England, and Ireland (Greer & Keohane, 2004; Greer, Keohane, & Healey, 2002). Thus, we believe that the evidence is robust and we hope that it can be useful to behavior analysts, neuroscientists, and linguists interested in a thorough analysis of the evolution of verbal behavior in children's development.
We have also speculated on the cultural evolution of verbal functions for our species relative to our proposed verbal developmental scheme (i.e., the role of cultural selection). Of course theories on the evolution of language are so extensive that some linguistic societies have banned their proliferation; yet, anthropologists and linguists are now suggesting there is new evidence to support the evolution of language (Holden, 2004). Some linguistic anthropologists may find the evolution of cultural selection of verbal operants and higher order verbal operants useful. It is even possible that the capacity for higher order operants and relational frames constitutes that which has been heretofore attributed to a universal grammar. Speaker and listener responses could have evolved from basic verbal operants to interlocking speaker and listener responding between individuals and within the skin of individuals (self-talk and naming)--an evolution made possible by our anatomical and physiological capacities to acquire higher order operants combined with cultural selection. Moreover, reading and writing functions also probably evolved as an extension of the basic speaker and listener functions; without them reading and writing would not have been possible, at least in the way it has evolved for the species.
"The human species, at its current level of evolution, is basically verbal, but it was not always so. ... A verbal behavior could have arisen from nonverbal sources and its transmission from generation to generation, would have been subject to influences which account for the multiplication of norms and controlling relations and the increasing effectiveness of verbal behavior as a whole." (Skinner, 1992, p.470)
Speaker/writer operants and listener/reader responses constitute an important if not the most important aspect of human behavior as adaptation to what is increasingly a verbal environment. Simply speaking, verbal behavior analysis is the most important subject of a science of behavior. We hope it is not too presumptuous of us to suggest that verbal behavior analysis can contribute to a developmental psychology that treats environmental contributions as seriously as it treats the non-environmental contributions. After all, biology has come to do so (Dugatkin, 1996; Robison, 2004).
While we can simulate human listener and human speaker functions with nonhuman species (Epstein, et al., 1980; Savage-Rumbaugh et al., 1978), the simulation of naming and other speaker-as own listener functions with nonhuman species remains to be demonstrated. Premack (2004) argues from the data that nonhumans lack the capacity for recursion. "Recursion makes it possible for words in a sentence to be widely separated yet be dependent on one another." (Premack, 2004, p. 320). We suggest that recursion may have made possible by the evolution of speaker as own listener capabilities in humans as a function of both neural capabilities and cultural selection. Premack (2204) also presents evidence that teaching is a strictly human endeavor. "Unlike imitation, in which the novice observes the expert, the teacher observes the novice--and not only observes, but judges and modifies." (Premack, 2004, p. 320; D. Premack & A. Premack, 2003). This describes the interaction we have characterized as what takes place in a learn unit. The conversational unit differs from the learn unit in that the conversational unit requires a reciprocal observation. Observational repertoires like those Premack (2004) described may be fundamental components that underlie and presage the evolution of nonverbal to verbal behavior.
While observation has been studied as a phenomenon, few if any studies have sought the possible environmental source for observational learning. We argue that observational learning differs from other indirect effects on behavior in that, observational learning results in the acquisition of new operants. Other types of observational effects on behavior result in the emission of operants that were already in the observer's repertoire. The kind of behavior change identified by Bandura (1986) was most likely of the latter sort since the presence or absence of the operants was not determined prior to the observational experience. Imitation results from a history that reinforces correspondence between the imitator and a model's behavior.
Some children do not have observational learning or have weak observational repertoires. In cases where observational learning has been missing we have induced it by providing certain experiences. It also may be possible that children do not have observational learning until they have certain experiences. In one study, we increased observational learning as a function of having individuals function as tutors using learn units that required tutors to reinforce or correct the responses of their tutees. It was the application of the learn unit per se, specifically the consequence component that produced the new observational repertoire (Greer, et al., 2004a). In another case with children who did not learn by observing peers, we taught them to monitor learn unit responses of their peers and observational learning emerged (Greer et al., 2004b; Pereira-Delgado, 2005).
This observing phenomenon involves a kind of consequent benefit similar to what the listener gains--specifically the extension of sensory reinforcement. Perhaps the teaching capacity involving reinforcement of the observed behavior of the learner is related to particular listener capabilities, while the recursion phenomenon is related to the interlocking speaker-listener capability. It is the interlocking speaker-listener-as-own-listener functions that make the more sophisticated milestones of verbal function possible. These functions make thinking, problem solving, and true social discourse possible . They also support the development of repertoires compellingly described in relational frame theory (Hayes, et al., 2000). Speech, and we argue, the compression of information through auditory stimuli in the human species, makes possible the more advanced speaker as own listener or textual responder as own listener and perhaps by extension the phenomenon of recursion. Regardless of whether our interpretations of the evidence is compelling, the evidence does reveal that a more complete picture of verbal behavior is evolving and that the role of the listener, and particularly the interrelationship between speaker and listener, is key to further advances in our understanding of verbal functions and their development within the individual.
Verbal Behavior Analysis, Comparative Psychology and the Neuroscience of Language
None of the work that we have described or related work in verbal behavior obviates the role of genetically evolved brain functions as neurology correlated with the presence of our suggested milestones of verbal behavior and the generative aspects of behavior cum language. The research in verbal behavior does not question, or eliminate, the importance or usefulness of neuropsychological researches. Alternately, the work in the neuroscience of language does not obviate the environmental verbal functions of language as behavior per se and as higher order operants that are increasingly identified in verbal behavior analysis. They are simply different sciences involved with different aspects of language. On the one hand, work in verbal behavior analysis is beginning to identify key environmental experiences in cultural selection and to suggest how neuropsychology can make the journey from MRI analyses to real verbal function--behaving with language outside of the skin. On the other hand, the work in the neurosciences of language is beginning to identify the behavior beneath the skin. It is compelling to consider the mutual benefit to obtaining a more comprehensive understanding of language by relating the efforts. Most importantly, combining the evidence and types of inquiry from both fields can help us teach a few more children to be truly verbal.
Behavior analysts have simulated language functions in non-humans (Epstein, et al. 1980; Savage-Rumbaugh, et al., 1978) and comparative psychologists have identified differences between the verbal behavior of primates and the verbal behavior of humans (Premack, 2004). Non-human species have not demonstrated a speaker-as-own-listener status. However, research in verbal behavior analysis has led to the acquisition of listener repertoires, speaker repertoires, speaker as own listener repertoires, and generative verbal behavior in humans who did not have those repertoires prior to special environmental experiences. Perhaps work in verbal behavior analysis with individuals who can acquire verbal repertoires as a result of special interventions provides a bridge. While our particular work is driven by applied concerns, it may have some relevance to the basic science of behavior, comparative psychology, and the neuroscience of language.
Reprinted with permission from the 2005 issue of Behavioral Development, 1, 31-48. The references were updated from the original publication, relevant quotations were added, and minor editorial changes were made.
We would like to dedicate this paper to the memory of B. F. Skinner who would have been 100 years old at its writing. His mentorship and encouragement to the first author served to motivate our efforts to master his complex book and engage in our experimental inquires. We are also indebted to others who kept verbal behavior alive in times when the critics were harsh and the audience was narrow. Among these are Jack Michael, Charles Catania, Ernest Vargas, Julie Vargas, Mark Sundberg, U. T. Place, Kurt Salzinger, Joe Spradlin, Joel Greenspoon, and the children we worked with who needed what verbal behavior could offer in order for them to become social and more cognitively capable. While the audience remains narrow, we are confident that the effects of research in verbal behavior will select out a larger audience.
Baer, D. M. (1970). An age -irrelevant concept of development. Merrill Palmer Quarterly, 16, 238-245.
Bandura, A, (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice-Hall.
Barnes-Holmes, D., Barnes-Holmes, Y., & Cullinan, V. (2001). Relational frame theory and Skinner's Verbal Behavior. The Behavior Analyst, 23, 69-84.
Becker, B.J. (1989). The effect of mands and tacts on conversational units and other verbal operants. (Doctoral dissertation, 1989, Columbia University). Abstract from: UMI Proquest Digital Dissertations [on-line]. Dissertations Abstracts Item: AAT 8913097.
Becker, W. (1992). Direct instruction: A twenty-year review. In R. West & L. Hamerlynck, Design for educational excellence: The legacy of B. F. Skinner (pp.71-112). Longmont CO, Sopris West.
Bijou, S., & Baer, D. M. (1978). A behavior analysis of child development. Englewood Cliffs, NJ: Prentice-Hall.
Carr, E.G. & Durand, V.M. (1985). Reducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18, 111-126.
Catania, A.C. (1998). Learning. Englewood Cliffs, N.J: Prentice-Hall.
Catania, A. C. (2001. Three types of selection and three centuries. Revista Internacional de Psicologia y Therapia Psicologia, I (1). 1-10.
Catania, A. C., Mathews, B. A., & Shimoff, E. H., (1990). Properties of rule governed behavior and their implications. In D. E. Blackman and H. Lejeune (Eds.) Behavior analysis in theory and practice, (pp. 215-230). Hillsdale NJ: Erlbaum.
Chavez-Brown, M. (2004). The effect of the acquisition of a generalized auditory word match-to-sample repertoire on the echoic repertoire under mand and tact conditions. Dissertation Abstracts International, 66(01), 138A. (UMI No. 3159725).
Chavez-Brown, M. & Greer, R. D. (2003, July) The effect of auditory matching on echoic responding. Paper Presented at the First European Association for Behavior Analysis in Parma Italy.
Chomsky, N. (1959). A review of B.F. Skinner's Verbal Behavior. Language, 35, 26-58.
Chomsky, N. & Place, U. (2000). The Chomsky-Place Correspondence 1993-1994, Edited with an introduction and suggested readings by Ted Schoneberger. The Analysis of Verbal Behavior, 17, 738.
Chu, H.C. (1998). A comparison of verbal behavior and social skills approaches for development of social interaction skills and concurrent reduction of aberrant behaviors of children with developmental disabilities in the context of matching theory. Dissertation Abstracts International, 59(06), 1974A. (UMI No. 9838900).
Culotta, E., & Hanson, B. (2004). First words. Science, 303, 1315
Decasper, A. J & Spence, M. J. (1987). Prenatal maternal speech influences ON newborn's perception of speech sounds. Infant Behavior and Development, 2, 133-150.
Deacon, T. (1997). The symbolic species: The co-evolution of language and the brain. New York: W. W. Norton & Company.
Donley, C. R., & Greer, R. D. (1993). Setting events controlling social verbal exchanges between students with developmental delays. Journal of Behavioral Education, 3(4), 387-401.
Dugatkin, L. A. (1996). The interface between culturally-based preference and genetic preference: Female mate choices in Pecilla Roticulata. Proceedings of the National Academy of Science, USA, 93, 2770-2773.
Emurian, H. H., Hu, X., Wang, J., & Durham, D. (2000). Learning JAVA: A programmed instruction approach using applets. Computers in Human Behavior, 16, 395-422.
Epstein, R, Lanza, R. P., & Skinner, B. F. (1980). Symbolic communication between two pigeons (Columbia livia domestica). Science, 207 (no. 4430), 543-545.
Fiorile, C.A. (2004). An experimental analysis of the transformation of stimulus function from speaker to listener to speaker repertoires. Dissertation Abstracts International, 66(01), 139A. (UMI No. 3159736).
Fiorile, C. A. & Greer, R. D. (2006). The induction of naming in children with no echoic-to-tact responses as a function of multiple exemplar instruction. Manuscript submitted.
Gautreaux, G., Keohane, D. D., & Greer, R. D. (2003, July). Transformation of production and selection functions in geometry as a function of multiple exemplar instruction. Paper presented at the First Congress of the European Association for Behavior Analysis, Parma, Italy.
Gewirtz, J. L., Baer, D. M., & Roth, C. L. (1958). A note on the similar effects of low social availability of an adult and brief social deprivation on young children's behavior. Child Development, 29, 149152.
Gilic, L. (2005). Development of naming in two -year old children. Unpublished doctoral dissertation, Columbia University.
Greenwood, C. R., Delquadri, J. C., & Hall, R. V. (1984). Opportunity to respond and student achievement. In W. L. Heward, T. E. Heron, J. Trapp-Porter, & Hill, D. S., Focus on behavior analysis in Education (pp. 58-88). Columbus OH: Charles Merrill.
Greer, R. D. (1987). A manual of teaching operations for verbal behavior. Yonkers, NY: CABAS and The Fred S. Keller School.
Greer, R. D. (2002. Designing teaching strategies: An applied behavior analysis systems approach. New York: Academic Press.
Greer, R. D., Chavez-Brown, M. Nirgudkar, A. S., Stolfi, L., & Rivera-Valdes, C. (2005a). Acquisition of fluent listener responses and the educational advancement of young children with autism and severe language delays. European Journal of Behavior Analysis, 6 (2), xxxxx-xxx.
Greer, R. D., Gifaldi, H., & Pereira, J. A. (2003, July). Effects of Writer Immersion on Functional Writing by Middle School Students. Paper presented at the First Congress of the European Association for Behavior Analysis, Parma, Italy.
Greer, R. D., & Keohane, D. (2004). A real science and technology of teaching. In J. Moran & R. Malott, (Eds.), Evidence-Based Educational Methods (pp. 23-46). New York: Elsevier/Academic Press.
Greer, R. D., Keohane, D. D. & Healey, O. (2002). Quality and applied behavior analysis. The Behavior Analyst Today, 3 (1), 2002. Retrieved December 20, 2002 from http://www.behavior-analystonline. com
Greer, R. D., Keohane, D., Meincke, K., Gautreaux, G., Pereira, J., Chavez-Brown, M., & Yuan, L. (2004a). Key components of effective tutoring. In D. J. Moran & R. W. Malott, (Eds.), Evidence-Based Educational Methods (pp. 295-334). New York: Elsevier/Academic Press.
Greer, R. D. & McCorkle, N. P. (2003). CABAS[R] Curriculum and Inventory of Repertoires for Children from Pre-School through Kindergarten, 3rd Edition. Yonkers, NY: CABAS[R]/Fred S. Keller School. (Publication for use in CABAS? Schools Only)
Greer, R.D., McCorkle. N. P., & Williams, G. (1989). A sustained analysis of the behaviors of schooling. Behavioral Residential Treatment, 4, 113-141.
Greer, R. D., & McDonough, S. (1999). Is the learn unit the fundamental measure of pedagogy? The Behavior Analyst, 20, 5-16.
Greer, R. D., Nirgudkar, A., & Park, H. (2003, May). The effect of multiple exemplar instruction on the transformation of mand and tact functions. Paper Presented at the International Conference of the Association for Behavior Analysis, San Francisco, CA.
Greer, R. D., Pereira, J. & Yuan, L. (2004b, August). The effects of teaching children to monitor learn unit responses on the acquisition of observational learning. Paper presented at the Second International Conference of the Association of Behavior Analysis, Campinas, Brazil.
Greer, R. D., & Ross, D. E. (2004). Research in the Induction and Expansion of Complex Verbal Behavior. Journal of Early Intensive Behavioral Interventions, 1.2, 141-165. Retrieved May 20, 2005 from http:/www.the-behavior-analyst-today.com
Greer, R.D., & Ross (in press). Verbal behavior analysis: Developing and expanding complex communication in children severe language delays. Boston: Allyn and Bacon.
Greer, R. D., Stolfi, L., Chavez-Brown, M., & Rivera-Valdez, C. (2005b). The emergence of the listener to speaker component of naming in children as a function of multiple exemplar instruction. The Analysis of Verbal Behavior, 21, 123-134.
Greer, R. D. & Yuan, L. (2004, August). Kids say the darnedest things. Paper presented at the International Conference of the Association for Behavior Analysis and the Brazil Association for Behavior Medicine and Therapy.
Greer, R. D. Yuan, L. & Gautreaux, G. (2005c). Novel dictation and intraverbal responses as a function of a multiple exemplar history. The Analysis of Verbal Behavior, 21, 99-116.
Hart, B. M., & Risley, T. R. (1975). Incidental teaching of language in the preschool. Journal of Applied Behavior Analysis, 8, 411-420.
Hart B. & Risely, T. R. (1996). Meaningful Differences in the Everyday Life of America's Children. NY: Paul Brookes.
Hayes, S. Barnes-Homes, D., & Roche, B. (2000). Relational frame theory: A Post-Skinnerian account of human language and cognition. New York: Kluwer/Academic Plenum.
Heagle, A. I., & Rehfeldt, R. A. (2006). Teaching perspective taking skills to typically developing children through derived relational responding. Journal of Early and Intensive Behavior Interventions, 3 (1), 1-34. Available online at http:/www.the-behavior-analyst-online.com
Holden, C. The origin of speech. Science, 303, 1316-1319.
Horne, P. J. & Lowe, C. F. (1996). On the origins of naming and other symbolic behavior. Journal of the Experimental Analysis of Behavior, 65, 185-241.
Jadlowski, S.M. (2000). The effects of a teacher editor, peer editing, and serving as a peer editor on elementary students' self-editing behavior. Dissertation Abstracts International, 61(05), 2796B, (UMI No. 9970212).
Karchmer, M.A. & Mitchell, R.E. (2003). Demographic and achievement characteristics of deaf and hard-of-hearing students. In Marschark, M & Spencer, P.E. (ed.) Deaf Studies, Language, and Education. Oxford, England: Oxford University Press.
Karmali, I. Greer, R. D., Nuzzolo-Gomez, R., Ross, D. E., & Rivera-Valdes, C. (2005). Reducing palilalia by presenting tact corrections to young children with autism, The Analysis of Verbal Behavior, 21, 145-154.
Keohane, D, & Greer, R. D. (2005). Teachers use of a verbally governed algorithm and student learning. Journal of Behavioral and Consultation Therapy, 1 (3), 249-259. Retrieved February 22, 2006 from http:/www.the-behavior-analyst-today.com
Keohane, D, Greer, R. D. & Ackerman, S. (2005a, November). Conditioned observation for visual stimuli and rate of learning. Paper presented at the third conference of the International Association for Behavior Analysis, Beijing, China.
Keohane, D., Greer, R. D., & Ackerman, S. (2005b, November). Training sameness across senses and accelerated learning. Paper presented at the third conference of the International Association for Behavior Analysis, Beijing, China.
Keohane, D. D., Greer, R. D., Mariano-Lapidus (2004, May). Derived suffixes as a function of a multiple exemplar instruction. Paper presented at the Annual Conference of the Association for Behavior Analysis, Boston, MA.
Lamarre, J. & Holland, J. (1985). The functional independence of mands and tacts. Journal of Experimental Analysis of Behavior, 43, 5-19.
Lodhi, S. & Greer, R.D. (1989). The speaker as listener. Journal of the Experimental Analysis of Behavior, 51, 353-360.
Longano, J. & Greer, R. D. (2006). The effects of a stimulus-stimulus pairing procedure on the acquisition of conditioned reinforcement for observing and manipulating stimuli by young children with autism. Journal of Early and Intensive Behavior Interventions, 3.1,135-150. Retrieved February 22 from http://www.behavior-analyst-online.com
Lovaas, O.I. (1977). The autistic child: Language development through behavior modification. New York: Irvington Publishers, Inc.
Lowe, C. F., Horne, P. J., Harris, D. S., & Randle, V. R.L. (2002). Naming and categorization in young children: Vocal tact training. Journal of the Experimental Analysis of Behavior, 78, 527-549.
MacCorquodale, K. (1970). On Chomsky's review of Skinner's Verbal Behavior. Journal of the Experimental Analysis of Behavior, 13, 83-99.
Madho, V. (1997). The effects of the responses of a reader on the writing effectiveness of children with developmental disorders. . (Doctoral dissertation, Columbia University, 1997). Abstract from: UMI Proquest Digital Dissertations [on-line]. Dissertations Abstracts Item: AAT 9809740.
Marion, C., Vause, T., Harapiak, S., Martin, G. L., Yu, T., Sakko, G., & Walters, K. L. (2003). The hierarchical relationship between several visual and auditory discriminations and three verbal operants among individuals with developmental disabilities. The Analysis of Verbal Behavior, 19, 91-106.
Marsico, M.J. (1998). Textual stimulus control of independent math performance and generalization to reading. Dissertation Abstracts International, 59 (01, 133A. (UMI No. 9822227).
McDuff, G. S., Krantz, P. J., McDuff, M. A., & McClannahan, L. E. (1988). Providing incidental teaching for autistic children: A rapid training procedure for therapists. Education and Treatment of Children, 11, 205-217.
Meincke-Mathews, K. (2005). Induction of metaphorical responses in middle school students as a function of multiple exemplar experiences. Dissertation Abstracts International, 66(05), 1716A. (UMI No. 3174851).
Meincke, K., Keohane, D. D., Gifaldi, H. & Greer (2003, July). Novel production of metaphors as a function of multiple exemplar instruction. Paper presented at the First European Association for Behavior Analysis Congress, Parma, Italy.
Michael, J. (1993). Establishing operations. The Behavior Analyst, 16, 191-206.
Michael, J. (1982). Skinner's elementary verbal relations: Some new categories. The Analysis of Verbal Behavior, 1,1-3.
Michael, J. (1984). Verbal behavior. Journal of the Experimental Analysis of Behavior, 42, 363-376.
Morris, E. K. (2002). Age irrelevant contributions to developmental science: In remembrance of Donald M. Baer. Behavioral Development Bulletin, 1, 52-54.
Nuzzolo-Gomez, R. (2002). The Effects of direct and observed supervisor learn units on the scientific tacts and instructional strategies of teachers. Dissertation Abstracts International, 63(03), 907A. (UMI No. 3048206).
Nuzzolo-Gomez, R. & Greer, R. D. (2004). Emergence of Untaught Mands or Tacts with Novel Adjective-Object Pairs as a Function of Instructional History. The Analysis of Verbal Behavior, 24, 30-47.
Park, H. L. (2005) Multiple exemplar instruction and transformation of stimulus function from auditory-visual matching to visual-visual matching. Dissertation Abstracts International, 66(05), 1715A. (UMI No. 3174834).
Pereira-Delgado, J. A. (2005). Effects of peer monitoring on the acquisition of observational learning. Unpublished doctoral dissertation, Columbia University.
Pennisi, E. (2004, February 24). The first language? Science, 303 (5662), 1319-1320.
Pinker, S. (1999). Words and rules. New York: Perennial.
Pistoljevic, N. and Greer, R. D. (2006). The Effects of Daily Intensive Tact Instruction on Preschool Students' Emission of Pure Tacts and Mands in Non-Instructional Setting. Journal of Early and Intensive Behavioral Interventions, 103-120. Available online at http://www.behavior-analystonline. org
Premack, D. (2004, January 16). Is language key to human intelligence? Science, 303, 318-320.
Premack, D. & Premack, A. (2003). Original intelligence. New York: McGraw-Hill.
Reilly-Lawson, T. & Greer, R. D. (2006). Teaching the function of writing to middle school students with academic delays. Journal of Early and Intensive Behavior Interventions, 3.1,135-150. Available online at http://www.behavior-analyst-online.org
Robinson, G. E. (2004, April). Beyond nature and nurture. Science, 304, 397-399.
Rosales-Ruiz, J. & Baer, D. M. (1996). A behavior analytic view of development pp. 155-180. In S. M. Bijou, New Directions in Behavior Development. Reno, NV, Context Press.
Ross, D. E. & Greer, R. D. (2003). Generalized imitation and the mand: Inducing first instances of speech in young children with autism. Research in Developmental Disabilities, 24, 58-74.
Ross, D.E., D. E., Nuzzolo, R., Stolfi, L., & Natarelli, S. (2006). Effects of speaker immersion on the spontaneous speaker behavior of preschool children with communication delays, Journal of Early and Intensive Behavior Interventions, 3.1,135-150. Available online at http://www.behavioranalystonline.com
Savage-Rumbaugh, E. S., Rumbaugh, D. M. & Boysen, S. (1978). Science, 201, 64-66.
Schauffler, G. and Greer, R. D. (2006). The Effects of Intensive Tact Instruction on Audience-Accurate Tacts and Conversational Units and Conversational Units. Journal of Early and Intensive Behavioral Interventions, 120-132. Available online at http://www.behavior-analyst-online.com
Schwartz, B.S. (1994). A comparison of establishing operations for teaching mands. Dissertation Abstracts International, 55(04), 932A. (UMI No. 9424540).
Selinski, J, Greer, R.D., & Lodhi, S. (1991). A functional analysis of the Comprehensive Application of Behavior Analysis to Schooling. Journal of Applied Behavior Analysis, 24, 108-118.
Sidman, M. (1994). Equivalence relations and behavior: A research story. Boston, MA: Authors Cooperative.
Skinner, B. F. (1989). The behavior of the listener. In S. C. Hayes (Ed.), Rule-governed behavior: Cognition, contingencies and instructional control (85-96). New York: Plenum.
Skinner, B.F. (1957, 1992). Verbal Behavior. Acton, MA: Copley Publishing Group and the B. F. Skinner Foundation.
Speckman, J. (2005). Multiple exemplar instruction and the emergence of generative production of suffixes as autoclitic frames. Dissertation Abstracts International, 66(01), 83A. (UMI No. 3159757).
Sundberg, M. L., Loeb, M., Hale, L., & Eighenheer (2001/2002). Contriving establishing operations for teaching mands for information. The Analysis of Verbal Behavior, 18, 15-30.
Sundberg, M.L., Michael, J., Partington, J.W., & Sundberg, C.A. (1996). The role of automatic reinforcement in early language acquisition. The Analysis of Verbal Behavior, 13, 21-37.
Sundberg, M.L. & Partington, J.W. (1998). Teaching language to children with autism or other developmental disabilities. Pleasant Hill CA: Behavior Analysts, Inc.
Tsai, H., & Greer, R. D. (2006). Conditioned preference for books and accelerated acquisition of textual responding by preschool children. Journal of Early and Intensive Behavior Interventions, 3.1, 3561. Available online at http:/www.the-behavior-analyst.com.
Tsiouri, I., & Greer, R. D. (2003). Inducing vocal verbal behavior through rapid motor imitation training in young children with language delays. Journal of Behavioral Education, 12, 185-206.
Twyman, J.S (1996a). An analysis of functional independence within and between secondary verbal operants. Dissertation Abstracts International, 57(05), 2022A. (UMI No. 9631793).
Twyman, J. (1996b). The functional independence of impure mands and tacts of abstract stimulus properties. The Analysis of Verbal Behavior, 13, 1-19.
Vargas, E.A. (1982). Intraverbal behavior: The codic, duplic, and sequelic subtypes. The Analysis of Verbal Behavior, 1, 5-7.
Williams, G. & Greer, R.D. (1993). A comparison of verbal-behavior and linguistic -communication curricula for training developmentally delayed adolescents to acquire and maintain vocal speech. Behaviorology, 1, 31-46.
Yoon, S.Y. (1998). Effects of an adult's vocal sound paired with a reinforcing event on the subsequent acquisition of mand functions. Dissertation Abstracts International, 59(07), 2338A. (UMI No. 9839031).
Author Contact Information
R. Douglas Greer, Ph.D., SBA, SRS
Graduate School of Arts & Sciences and Columbia University Teachers College
Box 76 Teachers College Columbia University
525 West 120th
New York NY, 10027
Dolleen-Day Keohane, Ph.D., SBA, Asst.RS
CABAS Schools and Columbia University Teachers College
2728 Henry Hudson Parkway
Riverdale, NY 10463
(1.) For information on and the evidence base for teaching as a science in CABAS schools and the CABAS[R] System see Greer (2002), Greer, Keohane, & Healy (2002), Selinski, Greer, & Lodhi (1991), Greer, McCorkle, & Williams, 1989, and http:/www.cabas.com. The findings of the research we describe have been replicated extensively with children and adolescents in CABAS[R] Schools in the USA, Ireland, Argentina and England and we believe they are robust. A book that describes the verbal behavior research and procedures in detail is in progress for publication in 2006 (Greer & Ross, in progress).
(2.) We chose the term listener emersion because it seemed particularly appropriate. The Oxford English Dictionary 2nd Edition, Volume V describes one usage of the term emersion as follows, "The action of coming out or issuing (from concealment or confinement). Somewhat rare." (OED, p. 177) Thus, once a child has acquired the listener repertoire, the child may be said to have come out of confinement to a pre-listener status. They have acquired an essential component of what is necessary to progress along the verbal behavior continuum--a verbal behavior development cusp.
(3.) It would seem that a certain history must transpire in order for a point-to-point correspondence between a word spoken by a parent and the repetition of the word by a child to qualify as an echoic operant rather than parroting. The child needs to say the word under the relevant deprivation conditions associated with the mand or the tact and then have that echoic evolve into either a mand or a tact. Once at least one of these events transpires, the parroting can move to an echoic. While more sophisticated operants and higher order operants or relational frames are basic to many sophisticated aspects of verbal behavior, the move from parroting is probably just as complex. The acquisition of echoing is the fundamental speech component of verbal functioning. One wonders how long, and under what conditions, it took for the echoic repertoire to evolve in our species. To evoke true echoics in children who have never spoken is probably one of the major accomplishments of the behavioral sciences. Indeed, the procedures we now use in verbal behavior analysis to induce first instances of vocal verbal operants have never been tried with primates, nor has the procedure to induce parroting. However, procedures for inducing parroting and echoics and other first instances of vocal verbal behavior have been successful in developing functional vocal verbal behavior in individuals who probably would have never spoken without these procedures. Amazing! There are even more fundamental components underlying even these response capabilities and aspects of observation show rich potential (Premack, 2004).
(4.) We use the term natural fracture to differentiate numerically scaled hypothetical relations from relations that are absolute natural events as in the determination of geological time by the identification of strata. To further illustrate our point, "receptive speech" is a hypothetical construct based on an analogy made between the computer "receiving inputs" to auditory speech events. It is an analogy, not a behavior or response class. Measures of receptive behavior are scaled measures tied to that analogy, as in test scores on "receptive" speech. However, listener behavior is composed of actual natural fractures (i.e., the child does or does not respond to spoken speech by another). In still another example, operants are natural fractures, whereas an IQ is a scaled measure of a hypothetical construct. Moreover, acquisitions of higher order operants such as the acquisition of joint stimulus control for spelling are also natural fractures.
Table 1. Evolution of Verbal Milestones and Independence Verbal Milestones Effects on Independent Functioning 1) Pre Listener Humans without listener repertoires are entirely Status dependent on others for their lives. Interdependency is not possible. Entrance to the social community is not possible. 2) Listener Status Humans with basic listener literacy can perform verbally governed behavior (e.g., come here, stop, eat). They can comply with instructions, track tasks (e.g., do this, now do this), and avoid deleterious consequences while gaining habilitative responses. The individual is still dependent, but direct physical or visual contact can be replaced somewhat by indirect verbal governance. Contributions to the well being of society become possible since some interdependency is feasible and the child enters the social community. 3) Speaker Status Humans who are speakers and who are in the in the presence of a listener can govern consequences in their environment by using another individual to mediate the contingencies (e.g., eat now, toilet, coat, help). They emit mands and tacts and relevant autoclitics to govern others. This is a significant step towards controlling the contingencies by the speaker. The culture benefits proportionately too and the capacity to be part of the social community is greatly expanded. 4) Speaker Listener a) Sequelics. Humans with this repertoire can Exchanges with responds as a listener-speaker to intraverbals, Others (Sequelics including impure tacts and impure mands. and Conversational Individuals can respond to questions for mand or Units) tact functions or to intraverbals that do not have mand or tact functions. The individual can respond as a speaker to verbal antecedents and can answer the queries of others such as, "what hurts?" "What do you want?" "What's that?" "What do you see, hear or feel?" One is reinforced as a listener with the effects of the speaker response. b) Conversational Units. Humans with this repertoire carry on conversational units in which they are reinforced as both speaker and listener. The individual engages in interlocking verbal operants of speaker and listener. The individual is reinforced both as a listener for sensory extensions, and also as a speaker in the effects speaking has on having a listener mediate the environment for the speaker. 5) Speaker as Own a) Say and Do. Individuals with this repertoire Listener Status can function as a listener to their own verbal Say Do behavior (e.g., first I do this, then I do Conversational that), reconstructing the verbal behavior given Units by another or eventually constructing verbal Naming speaker-listener behavior). At this stage, the person achieves significant independence. The level of independence is dependent on the level of the person's listener and speaker sophistication. b) Self-talk. When a human functions as a reinforced listener and speaker within the same skin they have one of the repertoires of speaker-as-own-listener. The early evidence of this function is self-talk; young children emit such repertoires when playing with toys, for example (Lodhi & Greer, 1990). c) Naming. When an individual hears a speaker's vocal term for a nonverbal stimulus as a listener and can use it both as a speaker and listener without direct instruction, the individual has another repertoire of speaker as own listener. This stage provides the means to expand verbal forms and functions through incidental exposure. 6) Reader Status Humans who have reading repertoires can supply useful, entertaining, and necessary responses to setting events and environmental contingencies that are obtainable by written text. The reader may use the verbal material without the time constraints controlling the speaker-listener relationship. The advice of the writer is under greater reader control than the advice of a speaker for a listener; that is, one is not limited by time or distance. Advice is accessible as needed independent of the presence of a speaker. 7) Writer Status A competent writer may control environmental contingencies through the mediation of a reader across seconds or centuries in the immediate vicinity of a reader on a remote continent. This stage represents an expansion of the speaker repertoires such that a listener need not be present at the time or at the same location as the writer. The writer affects the behavior of a reader. 8) Writer as Own As writers increase their ability to read their Reader: The Self- own writing from the perspective of the eventual Editing Status audience, writers grow increasingly independent of frequent reliance on prosthetic audiences (e.g., teachers, supervisors, colleagues). A more finished and more effective behavior-evoking repertoire provides the writer with wide-ranging control over environmental contingencies such that time and distance can be virtually eliminated. Writing can be geared to affect different audiences without immediate responses from the target audience 9) Verbal A sophisticated self-editor under the verbal Mediation for expertise associated with formal approaches to Solving Problems: problem solving (e.g., methods of science, logic, authority) can solve complex problems in progressively independent fashion under the control of verbal stimuli (spoken or written). The characterization of the problem is done with precise verbal descriptions. The verbal descriptions occasion other verbal behavior that can in turn direct the action of the person to solve the particular problem. A particular verbal community (i.e., a discipline) is based on verbal expertise and modes of inquiry are made possible. Table 2. Verbal Milestones and Components Mile- Components (Does the Child Have These Capabilities?) stones Pre- * Conditioned reinforcement for voices (voices of others listener controls prolonged auditory observation and can set the stage for visual or other sensory discriminations) (Decasper & Spence, 1987) * Visual tracking (visual stimuli control prolonged observation) (Keohane, Greer, & Ackerman, 2005a) * Capacity for "sameness" across senses (multiple exemplar experiences across matching across olfactory, auditory, visual, gustatory, tactile results in capacity for sameness across senses) (Keohane, Greer, & Ackerman, 2005b) * Basic compliance based on visual contexts and the teacher or parent as a source of reinforcement (The child need not be under any verbal control.) Listener * Discrimination between words and sounds that are not words (Conditioned reinforcement for voices occasions further distinctions for auditory vocal stimuli) * Auditory matching of certain words (as a selection/listener response) (Chavez-Brown, 2005; Greer & Chavez-Brown, 2003) * Generalized auditory matching of words (as a selection/listener response) (Chavez-Brown, 2005) * Basic listener literacy with non-speaker responses (Greer, Chavez-Brown, Nirgudkar, Stolfi, & Rivera-Valdes, 2005) * Visual discrimination instruction to occasion opportunities for instruction in naming (Greer & Ross, in press) * Naming (Greer, Stolfi, Chavez-Brown, & Rivera-Valdes, 2005) * Observational naming and observational learning prerequisites (Greer, Keohane, Meincke, Gautreaux, Pereira, Chavez-Brown, & Yuan, 2004) * Reinforcement as a listener (A listener is reinforced by the effect the speaker has on extending the listener's sensory experience; the listener avoids deleterious consequences and obtains vicarious sensory reinforcement.) (Donley & Greer, 1993) * Listening to one's own speaking (the listener is speaker) (Lodhi & Greer, 1989) * Listening to one's own textual responses in joining print to the naming relation (Park, 2005) * Listening and changing perspectives: Mine, yours, here, there, empathy (extension of listener reinforcement joins speaker) (Heagle & Rehfeldt, 2006) Speaker * Vocalizations * Parroting (Pre-echoic vocalizations with point -to-point correspondence, here-say joins see-do as a higher order operant), auditory matching as a production response (Sundberg, Michael, Partington, & Sundberg, 1996) * Echoics that occur when see-do (imitation) joins hear-say (echoic) as a higher order duplic operant (Ross & Greer, 2003; Tsiouri & Greer, 2003) * [Faulty echoics of echolalia and palilalia related to faulty stimulus control or establishing operation control] (Karmali, Greer, Nuzzolo-Gomez, Ross, & Rivera-Valdes, 2005) * Basic Echoic-to-mand function (a consequence is specified in and out of sight, here-say attains function for a few verbalizations leading to rapid expansion of echoics for functions mediated by a listener) (Ross & Greer, 2003; Yoon, 1996) * Echoic-to-tact function (generalized reinforcement control, the child must have conditioned reinforcement for social attention) (Tsiouri & Greer, 2003) * Mand and tacts and related autoclitics are independent (learning a form in one function does not result in use in another without direct instruction) (Twyman, 1996a, 1996b) * Mands and tacts with basic adjective-object acquire autoclitic functions (a response learned in one function results in usage in another under the control of the relevant establishing operation) (Nuzzolo-Gomez & Greer, 2005). This Transformation of establishing operations across mands and tacts replicated by Greer, Nirgudkar, & Park (2003) * Impure mands (mands under multiple control--deprivation plus verbal stimuli of others, visual, olfactory, tactile, gustatory stimuli) (Carr & Durand, 1985) * Impure tacts (tacts under multiple controls--deprivation of generalized reinforcers plus verbal stimuli of others, visual, olfactory, tactile, gustatory stimuli) (Tsiouri & Greer, 2003) * Tacts and mands emerging from incidental experience (naming and the speaker repertoires) (Fiorile, 2004; Fiorile & Greer, 2006; Greer, et al, 2005b; Gilic, 2005) * Comparatives: smaller/larger, shorter/longer, taller/shorter, warmer/colder in mand and tact functions as generative function (Speckman, 2005) * Generative tense usage (Greer & Yuan, 2004) * "Wh" questions in mand and tact function (i.e., what, who, why, where, when, which) (Pistoljevic & Greer, 2006) * Expansion of tact repertoires resulting in greater "spontaneous" speech (Pistoljevic & Greer, 2006; Schauffler & Greer, 2006) * Speaker Listener Exchanges with Others: Does the Child Have These Capabilities? * Sequelics as speaker (Becker, 1989) * Sequelics as listener-speaker (Becker, 1989; Donley & Greer, 1993) * Conversational units (reciprocal speaker and listener control) (Donley & Greer, 1993) Speaker as * Basic naming from the speaker perspective (learns tact Own and has listener response) (Fiorile & Greer, 2006; Listener Horne & Lowe, 1996) * Observational naming from the speaker perspective (hears others learn tact and has tact) (Fiorile & Greer, 2006; Greer, et al., 2004b) * Verbal governance of own speaker responses (say and do correspondence as extension of listener literacy for correspondence for what others say and nonverbal correspondence that is reinforced) (Rosales-Ruiz & Baer, 1996) * Conversational units in self-talk (listener and speaker functions within one's own skin in mutually reinforcing exchanges) (Lodhi & Greer 1989) Early * Conditioned reinforcement for observing books (Tsai & Reader Greer, 2006) * Textual responses: see word-say word at adequate rate improved by prior conditioning of print stimuli as conditioned reinforcement for observing (Tsai & Greer, 2006) * Match printed word, spoken word by others and self and printed word, spoken word and picture/object, printed word and picture/action (Park, 2005) * Responds as listener to own textual responding (vocal verbalization results in "comprehension" if the verbalizations are in the tact repertoire, e.g., hearing tact occasions match of speech with nonverbal stimuli) Writer * Effortless component motor skills of printing or typing (see-write as extension of see-do) * Acquisition of joint stimulus control across written and spoken responding (learning one response either vocal or written results in the other) (Greer, Yuan, & Gautreaux, 2005) * Writer affects the behavior of a reader for technical functions (mand, tact, autoclitic functions) (Reilly-Lawson & Greer, 2006) * Transformation of stimulus function for metaphoric functions (word used metaphorically such as in, "she is sharp as a pin") (Meincke-Mathews, 2005; Meincke, Greer, Keohane & Mariano-Lapidus, 2003) * Writes to affect the emotions of a reader for aesthetic functions (mand, tact, autoclitic functions as well as simile and metaphor for prose, poetry, and drama and meter and rhyme scheme for poetry) Writer as * Is verbally governed by own writing for revision Own functions (finds discrepancies between what she reads Reader and what she has written, writer and reader in the same skin) (Madho, 1997; Reilly-Lawson & Greer, 2006) * Verbally governs a technical audience by reading what is written as would the target audience (editing without assistance from others, acquire listener function of target audience requiring joint stimulus control between the writer and the listener audience) (Reilly-Lawson & Greer, 2006) * Verbally governs an aesthetic audience as a function of reading what is written as would the target audience (editing without assistance from others, acquire aesthetic listener function of target audience with tolerance for ambiguity) (Meincke-Mathews, 2005) Verbal * (Is verbally governed by print to perform simple Mediation operations (verbal stimuli control operations) for (Marsico, 1998) Problem * Is verbally governed by print to learn new stimulus Solving control and multiple step operations (the characterization of the problem is done with precise verbal descriptions). The verbal descriptions occasion other verbal behavior that can in turn direct the action of the person to solve the particular problem (Keohane & Greer, 2005). A particular verbal community, or discipline, is based on verbal expertise tied to the environment and modes of inquiry are made possible.)