In what way do people act directly and indirectly on the environment?
We operate directly on the physical environment in mechanical ways. Pushing, pulling, placing objects, etc. The effects are “automatic” and tied to physical relations between the response and environment. We act indirectly on the environment with verbal behavior. Here, the reinforcers are mediated by other people, but can be no less inevitable than direct effects.
How is the term “verbal behavior” an advantage over other terms such as “speech” and “language?”
Speech: it emphasizes vocal behavior. Hard to apply to other forms such as written or sign language.
Language: Often refers to the practices of a linguistic community, and not the actions of an individual.
Verbal behavior: It tends to emphasize the behavior of an individual, and lends itself well to identifying controlling variables. Moreover, it does not have a long history in traditional explanations.
What is the total verbal episode?
It is an account of the speaker and listener behavior.
On page 3 (top), Skinner is talking about “recent advances” in the analysis of behavior. To what is he referring?
His work, and the work of others, in EAB. JEAB was published in ’54, and there was a small group of people dedicated to this field.
Does Skinner see VB as fundamentally different from other human behavior?
No. If fact, this is an important and essential point of this book. Specifically, we can apply the techniques of EAB to verbal behavior.
What is Skinner’s view of “understanding?”
We understand verbal behavior to the extent that we can predict it, and eventually, control it by manipulating variables.
What have been the shortcomings of linguistics, syntactical, and semantic studies?
They have not developed the techniques needed for a causal analysis.
A formulation appropriate to all fields has not emerged. (i.e., to any field of human endeavor – see S&HB for the same point)
What characteristics of verbal behavior make it a good object of study?
Lots of it!
Facts are substantial and can be agreed on
Development of writing repertoires has provided a measurement system that is convenient and precise
What, then, is lacking?
A functional analysis of vb.
How is it that the concept of “ideas” has made it tempting to look inside the organism? What is the problem with this?
VB is often considered to be an expression of “ideas.” The ideas gave rise to the vb, and was thought to be the cause of it.
Problem: ideas can not be observed independently. In effect, they are explanatory fictions, as their evidence of existence is in the behavior they are said to cause.
Skinner also talks about “images” that are said to give rise to vb. Taken together, and including other such explicanda, these constitute explanatory fictions. What is the problem herein with these?
They “allay” curiosity and bring inquiry to an end.
What is the unfortunate consequence of the doctrine of ideas?
That vb has an independent existence apart from other behavior. Words are not the expression of ideas, and we do not “use” them in this way.
Skinner also tackles “meaning.” In some cases, meaning seems straightforward and sensible. What is this? When does it not?
When the word refers to a physical object, it is easy to describe its meaning as the thing referred to. But when there is no physical referent, then it becomes more difficult. It becomes even more difficult in sentences.
One of the techniques that Skinner discusses is “paraphrasing” as a way to explain the meaning of a word. What is the disadvantage?
We have not accounted for the word by paraphrasing.
Skinner, on page 10, is giving us a preview of what is to come. He will talk about a descriptive taxonomy of vb, and some explanations of the various forms. Multiple causation will be discussed as it was in S&HB. Also, what happens when the speaker and listener are the same person? Skinner will address this topic by presenting an interpretation of “thinking.”
Check out the important features of this analysis. What are they?
Looks at the behavior of individuals
Little use of experimental results
No survey of the “literature”
The book represents an extension of well-known facts about behavior to vb. Thus, it is an exercise in “interpretation”, not extrapolating experimental results.
Appealing to conditions and behavior that are measurable.
Skinner talks about “meaning” in the bottom of page 13, and top of 14. Explain in your own words what “meaning” is.
It is not a characteristic of a word, or a dependent variable, but a characteristic of the causes of the word, or the independent variable.
Skinner defines verbal behavior as that which is reinforced through the mediation of others. Is it just vocal behavior? Can behavior have verbal and nonverbal characteristics?
Verbal behavior is not just vocal, as other forms can have mediated reinforcement (signs, notes, pointing, gestures). Yes, behavior can have both characteristics. For example, behavior may have reinforcing effects on the physical environment, but also generate reinforcers produced by others.
On pages 15-16, Skinner talks about different kinds of records of vocal behavior. What are they?
EMG-like record of muscles involved
Acoustic record that shows the various pitches of sounds emitted
Phonetic record – speech broken into constituent sounds (NOTE: such a record makes no assumptions about functional significance of the phonics)
Discuss the main point regarding “single instance” vs “class.”
In the analysis of behavior, we can describe single instances, but when predicting behavior, we are really predicting classes of behavior that have a particular property. For example, if we predict that a person will open the door, we are less concerned with any topography, and more concerned with behavior that results in opening the door. Likewise, in vb we are interested not in the musculature, but in sound patterns that produce a given effect on the verbal community.
Discuss Skinner’s view of the “speech-sound” or phonetics issue.
Speech-sounds are not really the important issue in understanding vb. Indeed, if vocal behavior was not present, and we used non-vocal communication, the problems of vb would still remain.
What is “hypostasis?”
If someone asks “What did Shaun say today?” and you respond with “He said he will be here at 5 pm”, this is a reconstruction of the behavior of interest, and is unusual in science. We typically do not report behavior by imitating it. In a sense, the reconstruction is the “name” of the response (not the response itself, of course). This reconstruction is hypostatis, and the names of the responses are acoustically similar to the responses themselves. Moreover, the variables responsible for the reconstruction are not necessarily the same as those responsible for the original statement.
What is the difference between a phonetic report and direct quotation?
The former requires only a small repertoire to transcribe speech, but thousands of different “words.” Moreover, a direct quotation usually involves inferring something about the causal variables or the effects on a listener.
What is the difference between “response” and “operant?”
Response is a single instance and there is no requirement of controlling variables. An operant is a class of responses and requires a specification of controlling variables (at least one is required: its effect on the environment).
So, what is the size of the unit of VB? What is a repertoire?
In linguistics, the unit may be the words and morphemes, or phrases, clauses, sentences, etc. Any of these may come under functional control and are therefore identifiable units.
A repertoire is a collection of potential operants of a speaker. It is different from a “vocabulary” as a repertoire involves operants that require controlling variables, not just a collection of topographies.
On page 22, Skinner launches into a discussion of “Probability of response.” What are some of the evidence of this concept?
Emission: Whether or not the response is emitted. Strength is heightened when the vb is emitted under unusual, or inappropriate situations. For example, talking work at a party or sporting event, projective tests (see S&HB), verbal summator.
Energy level: An energetic “NO” may indicate strength of the vb. Pitch level may also vary with energy level (as db increases, pitch may also).
Speed: This refers to how quickly the various parts of the vb follow each other, or the latency of the vb following relevant variables.
Repetition: Whether or not the response is repeated after the initial emission. “NO, NO, NO” suggests a stronger operant than merely “NO.”
What are the limitations of the aforementioned indices of strength?
They do not always vary together. Exceptions to co-variation must be explained. Also, different values of a measure may not reflect strength, but form: DE-sert and de-SERT are different operants unrelated to the energy level. Thus, accents on different word parts make energy levels difficult. Moreover, these measures may be affected by special contingencies, such as speaking more energetically to deaf people, or repetition may be needed in loud rooms. In essence, there are contingencies that may require changes in energy, repetition, or speed that may have nothing to do with the actual “strength” of the operant.
The discussion of over-all frequency is difficult. First, what is meant by this measure? Then, what is a limitation?
Over-all frequency is rate of response. But, simply developing “word counts” can be misleading, as they may be different operant classes.
Discuss probability and predicting the single instance.
In the laboratory, we can easily see changes in rate as a function of IV s. But, we are often interested in the probability of a single forthcoming event. It is probably the case that such predictions will rely on rates. Indeed, in multiple determination of a behavior, we infer the contribution of each variable based on observations of rates.
On page 29, Skinner describes an experiment in which a bird learns to make a “figure 8.” What procedure is this?
NOTE: In this section, Skinner is discussing the well-known processes of reinforcement and extinction. On page 31, he also talks about discriminative control.
Talk about how behavior may be controlled by deprivation.
When behavior is reinforced with, say, food, then the behavior may be strengthened by food deprivation.
Give an example of how a deprivation may strengthen a class of behavior.
If a person has learned to get food by 1) going to the refrigerator 2) asking for food and 3) getting in the car and going to the store, then 1,2, and 3 will be strengthened by food deprivation.
Explain how the concept of “thirst” or “hunger” is unnecessary.
In predicting or controlling a given response, we do not change the motivational condition directly, but instead manipulate the controlling variables (e.g., deprivation) directly. Thus, the 2nd link “thirst” is irrelevant, as practical control will require we look at manipulable variables.
Deprivation Thirst “Water”
Explain the difference between escape and avoidance.
In escape, the behavior terminates or reduces some event. In avoidance, behavior terminates or reduces some 2nd event that precedes the first event.
VB that is reinforced by a characteristic consequence (often specified in the mand), and is therefore under functional control of deprivation/aversive stimulation. There is a unique relationship between the form of the mand and the particular consequence.
It is derived from terms such as “command” or “demand.”
Can you infer the function of VB by examining the form?
No. However, there are some forms that are often associated with particular variables such that the function may be inferred.
What does Skinner mean by the “total verbal episode?”
He is referring to the fact that we must account for the speaker and the listener behavior.
What is the difference between a request and command?
The maintaining variables for the listener differ. In a request, the listener complies for positive reinforcers. In a command, the listener complies to escape from conditioned aversive stimuli (threats) supplied by the speaker.
Skinner writes at some length about various classes of mands that differ with respect to functional control over the listener behavior. Give some examples.
Request and command: see above
Prayer: speaker generates an emotional predisposition in listener that strengthens behavior
Advice: Positive reinforcers not supplied by speaker
Permission: When listener is inclined to behave, but is exposed to threat. Speaker reduces it.
In classifying mands in terms of controlling variables of listener behavior, how is this different from traditional schemes?
Traditionally, such classifications often appeal to the “intention” of the speaker. But, Skinner shows that intention may be translated into contingencies of reinforcement of the listener behavior.
Skinner points out that frequent mands may cause the listener to “revolt.” How can this effect be mitigated?
Softening or concealing the mand character: “I’m thirsty” or “Would you mind getting me a drink?” instead of “water.”
Flattery or praise: “Get me a drink, my good man.”
What is a generalized mand?
A mand that is evoked by a variety of EOs (“please”).
What are some dynamic properties of mands? What may account for them? What about listener behavior?
Energy level, speed, intensity. These may be determined by EO strength, as well as history of reinforcement.
Listener behavior may be determined by EO properties of speaker behavior, as well as historical variables.
What is the effect of a long history of responding to mands?
Listeners may respond without any particular EO. In a sense, a generalized compliance class.
What is the significance of the example of the 2 math problems and the mand "do it on paper."
That listeners may not even be aware that they are responding to mands and other controlling variables. Indeed, this example shows the effects of multiple control, as words in the instructions strengthened addition in the first problem, and multiplication in the 2nd.
On page 44, Skinner revisits the "meaning" problem. What is the traditional view of meaning of mands in this passage? In Skinner's view?
The reinforcer that is specified in the mand. In Skinner's view, it is the reinforcer, as well as the EO. Thus, the "meaning" of any mand must specify the entire contingency: evocative variables and the reinforcers responsible for their effect.
What is Skinner's view of formal systems that embrace grammar and syntax?
Not in favor, because they do not suggest techniques that would explain the provenance of the vb.
Explain the point about children learning to cry for different reasons. What is the advantage of this account?
In newborns, crying may be an unlearned response to various EOs. However, crying may then be reinforced by events such as parent attention, and therefore be evoked by the relevant EO.
The advantage is that 1) terms can be defined by known experimental processes (conceptually systematic!) 2) it is consistent with explicanda of other verbal and nonverbal behavior
What determines the form of a mand? Are other stimuli relevant?
Contingencies of reinforcement provided by the listener and verbal community at-large determine the form of a mand, and what forms will be effective. However, other stimuli are important, as they may acquire discriminative control over the mand because they are present during reinforcement.
What, then, is the "extended mand?" Why do they occur?
It is a mand that occurs in the absence of the usual discriminative stimuli. For example, we may demand that our car "stop", that the door "open", etc. Or, we may issue mands to babies, dolls, or untrained animals. The can not possibly provide reinforcement, but they have characteristics that are similar to individuals who do provide reinforcement for mands. Thus, it is a stimulus induction, or generalization, process.
NOTE: because some of the usual controlling variables are absent, the response may be unusually weak or “whimsical” with a comment that indicates its lack of strength.
How could mands be superstitious?
Certain mands may be maintained by occasional reinforcers that have no mechanical or physical connection. In a sense, the reinforcers are “accidental” or delivered for reasons other than the occurrence of the mand.
Can we mand "emotional reactions?" (Cheer up! Dry your tears!)
In some cases, the mere form of the mand does not have the desired effects. However, there are other stimuli that may accompany that may have the desired effects. For example, "Dry your tears" may be accompanied by stroking or soothing tones of voice that may affect the listener.
What are magical mands? How does Skinner account for them?
It is a form of extended mand that has no history of reinforcement to explain its emission, or the emission of similar behavior.
Skinner accounts for them by appealing to a history of mands that forms a generalized response class that is maintained by "control over the environment." Or, there may be moments of "great stress" (or strong EOs) that may evoke the response; in a sense, the speaker may be merely describing the reinforcer appropriate to a given EO.
How does Skinner explain poetry that includes mands?
He discusses "poetic license" in which the writer experiences strong EOs, and emits mands in the form of lyric poems. The poetic community and readership maintain it, despite the absence of usual contingencies for mands.
Verbal Behavior Under the Control of Verbal Stimuli
Are prior stimuli irrelevant in the mand?
No. Mands are reinforced only when an audience is present, so discriminative control by the audience emerges.
What are the 3 levels of relative probabilities of the mand? Is “C” the same as “naming?”
Absence of listener (low probability)
Listener appears (higher probability)
Listener appears with desired item (highest probability)
No. Naming is more appropriate to vb that bears no relationship to a specific reinforcer, and is therefore independent of a EO.
What are the two ways of releasing vb from a specific EO?
Reinforce with a variety of reinforcers, and the response will exist in strength unless there is no EO
Use generalized conditioned reinforcement (approval, praise)