Verbal Behavior

Chapter 12

Verbal Behavior

Synopsis:

Verbal behavior focuses on the function of speaking, writing, reading, signing, and other forms of communication. Unlike many approaches that emphasize the structure of language, including grammar and syntax, the verbal-behavior approach views language as a product of contingencies of reinforcement in the environment and mediated by the behavior of other organisms. That is, verbal behavior only works through its effects on other people.The capacity for verbal behavior is a product of natural selection operating on the vocal and neural systems (including larynx, tongue, and neuromuscular systems) allowing for the productionof speech sounds, which comeunder operant control. In humans, the FOXP2 FoxP2 transcription factor (encoded by the FOXP2 FoxP2 gene) is implicated in enhanced motor control and learning of speech. The contingencies that regulate verbal behavior arise from the practices of people in the verbal community and determine the function and form of speaking (specific language spoken). Manding and tacting are two basicclasses of verbal operant behavior. Manding is a verbal operant class regulated by establishing operations (EOs) and maintained by specific reinforcement. Tacting is a verbal operant class regulated by nonverbal discriminative stimuli and maintained by generalized conditioned reinforcement. Other basic verbal operants include: (a) echoics, (b) textual relations, (c) intraverbals, and (d) autoclitics. The basic units of verbal behavior are combined to form higher-order classes, allowing for the greater novelty and spontaneity in a child’s language production,includinghigher-order verbal classes like naming. Through multiple- exemplar instruction (MEI) the child learns that listener and speaker responses go together, establishing the generalized verbal class of naming. As verbal repertoires increase across the lifespan, verbal behavior becomes more symbolic as shown byresearch on learning of stimulus equivalence and formation of thestimulus classes denoted as reflexivity, symmetry, and transitivity—basic logical relations of mathematics. Complex symbolic human behavior is built on the equivalence relation and other derived stimulus relations, and the neural basis of derived relations is an ongoing area of behavioral neuroscience. Stimulus equivalence training has been helpful to those who lack reading skills and in the development of educational curricula based on derived stimulus relations. Research shows that animals can be trained to pass tests of stimulus equivalence, but that these relationsare substantially extended in human behavior. Analysis of three-term contingencies of mother–-child interactions in natural speech supports the operant perspective on the early development of verbal behavior.A detailed analysis in the advanced section of this chapter depicts the social contingencies regulating the basic verbal relations of manding and tacting as outlined by Skinner (1957) in his book Verbal Behavior.

Study Questions:

What is language and verbal behavior? How does verbal behavior differ from language?
Identify developmental changes in the larynx and tongue of human children that facilitate production of the vowel sounds [i], [u], and [a].
How did Skinner describe the differences between the behavior of listener and speaker?
Describe the differences between the information-transmission view of language and the behavior analyst view.
Define the following concepts: (a) manding and (b) tacting.
Describe the procedure of the blocked-response conditioned establishing operation (CEO). How is this procedure used toin train manding relations?
With regards to the tacting relation, what is nonspecific reinforcement?
What did Carrol Carroll and Hesse’s (1987) research regarding the alternating training of tacting and manding suggest about the independence of these classes of behavior?
What is Skinner’s functional independence hypothesis?
Defineintraverbal behavior and provide a common example based on learning of addition in school.? How does anintraverbal response differ from anechoicresponse—? Ttry to use the learning of addition as an example throughout your answer?
Identify the two requirements for a response to be considered echoic behavior.? How is textual behavior different?
What is an autoclitic and what are the five categories of autoclitic relations?
According to applied behavior analysts, how should autoclitics be trained?
From what basic verbal behaviors does the naming relation develop?
Identify and define the three basic classes or typesof equivalence relations.
Identify five areas of the brain that neuroscientists have found to be involved in derived stimulus relations.
In Frank and Wasserman’s (2005) research on equivalence in pigeons, what training condition was necessary for pigeons to demonstrate evidence of symmetry?
How do behavior analysts explain evidence that reinforcement may not be necessary for the development of emergence of derived stimulus functions?
What did Moerk’s (1990) research on mother-–child communication suggest about the development of language usage in humans?
What is an interlocking contingency?Show how both mand and tact relations are part of interlocking contingencies.?

Study Questions (Answers):

Language is a poorly defined concept that can refer to linguistic habits, universal grammar, vocabulary, syntax, and meaning of communication. Verbal behavior deals with the performance of a speaker and the environmental conditions that establish and maintain such performance. Verbal behavior focuses on the function of what we say.
During development, the larynx descends into the adult position in the neck (C6), changes shape in ways that alter breathing and swallowing, and muscular control is extended within the larynx and pharynx. During this same time, the tongue descends into the throat (pharynx), assumes a posterior rounded contour, and the tongue’s oral and pharyngeal proportions become equal (1:1).
Skinner used the term rule-governed behavior to describe the behavior of the listener and verbal behavior to describe the behavior of the speaker. Rules are complex verbal discriminative stimuli for the listener, and the principles that govern stimulus control also regulate the rule-following behavior of the listener. Verbal behavior is shaped by subtle contingencies of reinforcement mediated bythe behavior of the listener.
Information-transmission assumes that words refer to things in the world and have the power to represent, communicate, and express the world as perceived by the speaker. The behavior analysis view is that the social environment shapes the way in which we use words. The way we talk and what we say are a function of social contingencies involving effects or consequences arranged by members of a verbal community.
Manding is a class of verbal operants whose form is regulated by establishing operations and maintained by specific reinforcement. Tacting is a class of verbal operants whose form is regulated by nonverbal discriminative stimuli and maintained by generalized conditioned reinforcement from the verbal community.
The blocked-response conditioned establishing operation (CEO) procedure involves temporarily withholding a condition, stimulus, or event that is necessary for the completion of a behavior pattern or chain. This procedure is often used to train manding relations by creating an establishing operation for requesting the withheld stimulus, event, or condition.
Nonspecific reinforcement refers to a situation where the reinforcer for the response exerts no stimulus control over the form of the next response (e.g., seeing a picture of an apple, saying “apple,”, and then receiving verbal praise).
Carrol Carroll and Hesse (1987) found that mand training facilitated the acquisition of tacting. This suggests that under some conditions, manding and tacting are not independent classes of behavior. This may be the case when parts of the response forms are shared.
Skinner’s functional independence hypothesis states that manding and tacting are separateresponse classes controlled by distinct contingencies. Due to the differences in controlling contingencies, it should be possible to train tacting and manding independently.
Intraverbal behavior is a class of verbal operants regulated by verbal discriminative stimuli. A common example of the intraverbal is learning addition in school.The teacher says “What is two plus one” and the student answers “three.”. An intraverbal does not involve a point-to-point correspondence between the verbal stimulus (2 + 1 =) and the response (writing the number 3). When point-to-point correspondence exists, this is termed echoic behavior (teacher says “three” and student says “three”).
Echoic behavior requires that the response has (a) point-to-point correspondence and (b) formal similarity, meaning that the behavior occurs in the same mode (auditory or visual) as the stimulus. Textual behavior requires a point-to-point correspondence, but not formal similarity.
The autoclitic is a form of verbal behavior that modifies the consequences produced by other verbal responses and is used in conjunction with, and controlled by, primary verbal units (mands, tacts, and intraverbals). The five categories of autoclitic relations are descriptive, qualifying, quantifying, manipulative, and relational.
Applied behavior analysts recommend that autoclitics be trained indirectly by focusing on a wide variety of primary verbal operants such as mands, tacts, echoics, and intraverbals to allow for “communication effectiveness” and contact with the natural contingencies of the verbal community.
Horne and Lowe (1996) stated that the naming relation (or generalized class of naming) arises from verbal contingencies that integrate the echoic and tacting response classes of the child as speaker with the conditional discrimination behavior of the child as listener.
The three basic classes or types of equivalence relations are: (1) symmetry— – when stimulus class A is shown to be interchangeable with stimulus class B;, (2) reflexivity or identity— – the A to A relation;, and (3) transitivity— – if A = B and B = C, then A = C.
Neuroscientists using fMRI technology have found the following areas to be involved in derived stimulus relations: frontal right-hemisphere (non-verbal areas), anterior hippocampus in the medial temporal lobe (memory areas), the parahippocampus (memory- encoding area), and the frontal and parietal lobe regions.
Frank and Wasserman (2005) found that symmetry was only demonstrated when the training condition included intermixed A–-A and B–-B identity training.
Evidence suggests that stimulus–-stimulus (S–-S) correlation, the degree to which one stimulus goes together with another, is the basic determinant of emergent relations, not reinforcement contingencies.
Moerk (1990) found that many mother–-child–-mother verbal sequences ended with maternal reinforcement in the form of feedback that confirmed the child’s verbal behavior. This finding indicates that three-term contingencies characterized many of the verbal episodes of early language imitation.
An interlocking contingency involves the intermingling of behavior chains during social interaction in which each person’s verbal responses function as discriminative stimuli and reinforcement for the other person’s behavior. In the mand relation, deprived of ketchup the speaker’s mand (“Ppass the ketchup”) functions as a discriminative stimulus for the listener’s behavior (passing the ketchup) and the listener’s response (passing the ketchup) functions as specific reinforcement of the speaker’s manding. In tacting, the verbal prompt (“What’s that?”) from the listener in the presence of the ketchup bottle (nonverbal SD)sets the occasion for the speaker’s verbal response (“Ketchup.”) that is followed by generalized reinforcement by the listener (“Thanks”).

Essay Questions (Student):

Describe the general findings of Lamarre and Holland’s (1985) and Petursdottir, Carr, and Micahel’sMichael’s (2005) tests of Skinner’s functional independence hypothesis. Discuss three possible reasons for the conflicting findings between these studies.

ANS: Lamarre and Holland (1985) demonstrated the functional independence of manding and tacting in children between 3 and 5 years old. The results clearly supported Skinner’s functional independence hypothesis.

Petursdottir, Carr, and Michael (2005) used preschool children to systematically replicate and extend the findings of Lamarre and Holland (1985) and found that following mand training, 4 out of 4 children reliably emitted tact responses on probe trials, but tact training produced unreliable effects on tests for mand responses.

One possibility is that the contingencies differed between the two studies. The earlier study by Lamare Lamarre and Holland (1985) required the children to emit verbal responses to an abstract stimulus property (location on the left or on the right), whereas the more recent study by Petursdottir et al. (2005) used a concrete stimulus (pieces of puzzles or cubes) to establish the requisite verbal behavior.

Another way to account for the differences relates to establishing operations. Lammare Lamarre and Holland did not use an explicit establishing operation (EO) for mand training, but the assembly-task study did—using concealed pieces of the task as a specific EOestablishing operation.

The reinforcement histories of the children also differed between the two studies. When children have a relevant history of manding and tacting outside the laboratory, as would be the case for verbal responses involving objects (pieces of puzzle), these response classes would show more cross-transfer in an experimental setting.

What is multiple- exemplar instruction (MEI) and why is it important to the development of the naming relation in children with language delays? How do outcomes using multiple exemplar instructionMEI compare to the use of single example exemplar instruction (SEI)?

ANS: A variety of research studies of children with language delays indicate that the naming relation arises from multiple exemplar instructions (MEI), involving rotation of the child’s listener and speaker responses during training. After training to establish basic verbal units and listening, MEI alternates among instructions to match, instructions to point, and instructions to tact arranged in different sequences. Results of various studies show that MEI training increased novel unreinforced listener and speaker components of naming, but that single exemplar instruction (SEI) failed to show the emergence of novel naming. The data suggest that rotation of the speaker–listener components found in MEI training is required for the acquisition of naming in children who lack a naming repertoire. Thus, through MEI the child learns that listener and speaker responses go together, establishing the generalized verbal class of naming. The speaker–listener repertoires remain independent with SEI training, and there is no evidence of generalized naming with these procedures.

Essay Questions (Instructor):

Describe how echoic relations developed at an early age may serve as the foundation of more complex verbal behavior.

ANS: Echoic behavior occurs at an early age in an infant’s acquisition of speech. The child who repeats “dada” or “mama” to the same words uttered by a parent is showing echoic operant behavior. In this situation, any product of behavior (sound pattern of child) that closely replicates the verbal stimulus (modeled sound pattern) is reinforced. The contingencies of echoic behavior are probably based on the matching of phonetic units. Catania indicates that the learning of echoic behavior begins with the basic units of speech called phonemes. Coordinated movements of the child’s larynx, tongue, and lips result in phonemes (e.g., “ma”), which replicate parts of adult speech (“ma ...… ma”). When articulations by the child correspond to those of the adult, the acoustical patterns also overlap. Adults who hear speech-relevant sounds (“ma”) often provide social consequences (e.g., tickling, tummy poking, or smiling) that are paired with these acoustical patterns. On this basis, the duplication of speech sounds itself comes to function as automatic reinforcement for speech-relevant articulations by the child. These units begin as phonemes (i.e., the smallest sound units to which listeners react), expand to words, and eventually may include full phrases and sentences.

Define the following elements of stimulus equivalence and give an example of each based on a child learning to read: identity, reflexivity, symmetry, and transitivity. Give an example of each.Describe the components of a complete experiment testing for stimulus equivalence relations.

ANS: Identity occurs when an organism selects stimulus class A when presented with stimulus class A (A = A; B = B). The child is shown a picture of a dog and points to the dog among animals in the book. Symmetry occurs when stimulus class A is shown to be interchangeable with stimulus class B (if A = B then B = A). The child points to a dogwhen shown the photograph of a dog and when asked to find “the dog” in the book.In this case, the picture and the word have anequivalent equivalence in function.Transitivity occurs when it the child is shown that if stimulus class A = stimulus class B and stimulus class B = stimulus class C, then stimulus class A = stimulus class C. The child points to a dog when shown the photo of a dog, points to the dog among animals in the book when asked to find “the dog,” and points to the dog in a group of animals when hearing a “barking” sound. The picture, word, and sound are shown to be equivalent in function (Picture of Dog = Word Dog = Sound of Dog).

A complete experiment for stimulus equivalence consists of both identity and symbolic matching procedures. In identity matching, the researcher presents a sample stimulus and two options. The procedure is repeated over multiple examples of sample and comparison options. The organism is reinforced for choosing the option that corresponds to the sample, establishing generalized matching-to-sample or generalized identity matching. Symbolic matching involves presenting one class of stimuli as the samples and another set of stimuli as the options. After the reinforced relations are trained, tests are made for each kind of equivalence relation. The question is whether reflexivity, symmetry, and transitivity occur without further training.

Videos:

B.F. Skinner— – Focus on vVerbal bBehavior

Manding— - Training

Tacting— - Training

Echoic relations

Intraverbal relations

Not as scary as you think: Sstimulus equivalence

Verbal bBehavior pPanel dDiscussion (complex verbal behavioris analyzed by several members of the panel)

Study Tips:

Verbal behavior is about the function of speaking, writing, signing, and other forms of communication. The goal is to analyze the environmental contingencies that shape and control verbal behavior of the speaker.
Manding is verbal behavior under the control of an establishing operation (EO) and specific reinforcement. Tacting is verbal behavior controlled by nonverbal discriminative stimuli and generalized conditioned reinforcement or nonspecific reinforcement.
Echoic behavior involves both a point-to-point correspondence and formal similarity. Textual relations involve a point-to-point correspondence, but not formal similarity.
Stimulus equivalence involves: reflexivity or identity matching (A = A, B = B), symmetry (A = B and B = A), and transitivity (A = B and B = C, then A = C).
Most human verbal interactions involve interlocking contingencies built on reciprocal exchanges of the basic verbal operants (manding, tacting, echoics, textual relations, autoclitics, and intraverbals).