This BCBA-D Rejects Verbal Behavior as a framework for language-based intervention.

Why? Because Verbal behavior is not language...

because language is not behavior.

Speaking is behavior. Writing is behavior. Speaking and writing have duration, intensity and frequency like all behavior. Uttering a promise is behavior. The promise itself is not behavior. So when someone says, "I promise to take you to dinner for your birthday"... uttering these words is behavior. But the promise itself is not behavior. It is the content of the utterance. Verbal behavior concerns only the utterance and what are purported to be 'controlling' events and is not concerned with what is spoken or written or that which which is communicated, whether a promise, a fact or an invitation is communicated. That which is communicated is communicated in language. What is expressed, whether a feeling thought, or an opinion is expressed in language. The utterance, the speech, the writing the gesture is behavior. Communication can't take place outside of the behavior which carries it, but what is communicated is not behavior. And despite a common meme that applied verbal behavior involves teaching children to communicate... it does not. It has nothing to do with communication.

Verbal Behavior and language are often conflated. Language, according to behavior analysts is reducible in terms of verbal behavior or reported to be 'language'.

Tina Sidener (2010) says,


“B. F. Skinner wrote a book called Verbal Behavior, in which he

introduced a controversial idea – that language is behavior!”

Skinner never said that.


He said, "Verbal behavior is behavior".

Many highly visible members of the behavior analytic community perpetuate this meme. Carbon talks about "behavioral categories of language". But this is confused. This is based on a false premise... language is not behavior, so there can be no "behavioral categories of language". Language is an anthropological phenomenon. There are no behavioral categories in anthropology. Anthropologists learn about culture through fieldwork and first hand observation. It is a descriptive endeavor. It is not an experimental / scientific enterprise.

This confused meme has been adopted by the larger "scientific" culture as well. On the website of the  American Psychological Association it is stated,


“Verbal behavior: The term is used by those who favor a behavioristic account of language”.

And in this frenzied 'ABA' industry, insurance companies insist on using assessment based on a verbal behavior scheme. This scheme is a taxonomic system. It answers "what something is", to what category something belongs, not what something is used for. This scheme lends itself easily to data collection. Practitioners count and report tacts, mands and intraverbals. Discrete stimuli are presented to which there are scorable responses. Children, in fact, do learn to do these kinds of things. And when they do it is exciting because before they learned to 'respond' in these ways, learning had been elusive. Graphs are clean. One behavior at a time can be accounted for (e.g., answering 'what color is it', 'what is a fork for', "where do you sleep", naming things etc.) But children are not learning a language. They are learning to emit discrete utterances. They are not inviting, promising, negotiating, explaining, cajoling etc.


Language is messier... tracking progress is eventually less discrete. Moves in language occur within a context, within activity. Doing more than one thing at time is often required... e.g., Asking a child to hang up a picture may require that they look for the proper tools, asking questions, seeking assistance, taking directions about where something goes, making adjustments along the way etc. But, the culture of insurance companies finds an AVB approach suitable to their needs. And, they have been sold on the 'science' of applied verbal behavior. But science is not called for in language learning.

In spite of its popularity and acceptance, and in spite of the conflated claims of venerated experts (I've heard some referred to as "legends"), such claims run counter to what Skinner actually says. Skinner was not attempting to account for language in behavioral terms or for language itself. He distinguishes language from verbal behavior when he says, "...language is a linguistic practice of the verbal community" (1957, p.2). Skinner was interested in the behavior of speakers... not cultural practices and not anthropology.


Language and verbal behavior are based on different conceptual schemes. Verbal behavior is based on operant behavior principles and behavioral philosophy. It is a deterministic, (quasi) scientific endeavor. As a 'scientific' enterprise, Skinner hoped to reduce the behavior of speakers to causal events... such that there is some force "x" acting on something else. He intended to sanitize utterances from intent, meaning and matters of 'mind'; to reduce this form of life to utterances devoid of meaning or intention. As such, the behavior of speakers is said to be caused by environmental events or more specifically is 'controlled' by them. There is no agent/person using words for a purpose. There is only the force of the environment 'causing' sounds to be produced by an unwitting organism. A 'stimulus - response' paradigm sits under the verbal behaviorist's yammer.

Let's consider an example. In this world view, the utterance of "Hi" is explained in terms of causal events. Thus, in a verbal behavior scheme, the reason"Hi" is uttered by someone upon first seeing someone else would be accounted for as either mand, tact, intraverbal or echoic relations. Quickly surveying which of these operant relations is at play, it's likely that behavior analysts would conclude that saying "Hi" is a mand (since it is not possible that it is a tact, intraverbal or echoic since there is no verbal antecedent stimulus or event or property that would evoke, "Hi"). But by saying that the utterance "Hi" is a mand relation means there is an EO or MO evoking this 'response'. Therefore, saying"Hi" indicates there is a state of deprivation operating...ostensibly there is deprivation of "hi" and consequently speakers name their own reinforcer, "hi". This is the 'science...silly as it is. And if you wish to hold to the scientific merits of AVB, try to verify or falsify the force of EOs or MOs. They can't be verified or falsified.They are only inferred. This is no different from saying someone "wants' something. No experimentation will reveal what sits behind saying "Hi".

But in language, what we say is informed by and part of cultural practice. This conceptual framework has nothing to do with causation or science. Thus, saying "Hi" when first seeing someone is a part of a greetings practice in a given culture. Similarly, if someone responds to "Hi",  by saying, "You look great!", "I've missed you", or "I'm so angry with you!", there is no need to concoct explanatory systems supporting the behaviorist's claim that all of these different responses were learned and emitted be'cause' of the presence of some controlling antecedent stimulus. Such responses are simply things deemed appropriate within a practice. There is no need to invoke 'controlling stimuli'. No forces are exerting themselves on persons ... compelling them to respond in determined ways. No explanation is required. No analysis needed. What is done in the practice is survey-able. No theory required.


As members of an American English speaking culture, teaching others language requires they learn the rules for the use of words (the meanings of the words; the words which express the concepts underlying them) and the many practice forms of the language ... the kinds of things done with words. Our parents, teachers, aunts and uncles taught use how to participate in this practice. None of them, at least when I was growing up, had advanced degrees in behavior analysis... they simply offered basic guidance in the practice... teaching us the meanings of words and how to put those words in to play within the many practice forms of the language (e.g. negotiating, refuting, telling, inviting, promising etc.)

What does this mean for intervention. It means we need to understand the premise of each... to be clear about what we are attempting to effect. That we are clear about the aims of each, for they are not the same. We can either teach verbal behavior or language, i.e., we are either indoctrinating others in a linguistic/cultural practice or teaching persons to be emitters of sounds. Teaching verbal behavior is not to teach language. One can't be used to teach the other or to account for the other. Doing so is like trying to teach someone to play chess according to the rules of checkers. Attempting to explain chess in terms of checkers is a recipe for confusion and incoherent ramblings.

The goals of "verbal behavior"as an applied intervention:


> That 'learners' learn to emit sounds which are 'evoked' by controlling verbal or nonverbal antecedent stimuli. For example, a person might emit the sounds 'red shirt' in the presence of someone wearing a red shirt. Behavior analysts refer to such utterances as 'tact relations'. If someone is also wearing a yellow hat, maybe that would evoke"yellow hat, red shirt" (Of course, we don't see people actually going around naming things all day even if such responses have been learned and remain strong...  why don't we see people doing this?) And a verbal antecedent, "Hi"uttered by a person in a red shirt and yellow hat, evokes a learned response, "Hi", or "Hello". This would have to be categorized as an echoic relation. (Absurd as it is to reduce this basic social practice to such terms... but there it is). How is this helpful? And, why shouldn't the 'listener' respond,"Hi. Red shirt yellow hat?" Why don't people talk this way? (Behavior analysts would say this is perfectly ok and would proclaim that 'joint control' is operating) How can such 'analyses" inform our efforts to teach children to learn what to do in a greetings games? Saying such things during a greeting would just be odd.


> That 'learners' accumulate, hundreds of thousands of verbal units across categories of verbal response classes i.e., mands, tacts, intraverbals, echoics, etc. (Partington and Sundberg, 1998). Such units are discrete and independent of each other. In order to facilitate learning, Sds are culled from lists, to which responses are taught. These responses are taught for their own sake.  Children learn to respond to arbitrary 'Sds' such as, "When do you go to sleep?". A response, "At night" is taught so to form an intraverbal unit. (Whether a child actually knows what 'night' means is not relevant). Of course, other responses can be taught to the same 'Sd'...but each intraverbal unit is independent such that separate learning or generalization takes place for each antecedent stimulus. 


Does this mean that all the words on this page are in response to some specific controlling stimulus or 'possibly' that response generalization has taken place? Does this mean also that every lecture ever given, every wish wished or promise promised was in response to some specific controlling stimulus? According to behavior analysis, the answer is "Yes"... every 'response' is a function of either direct or indirect learning to some specific controlling stimulus. If this is so, then saying that someone 'formulates a response', or 'thinks through what they are going to say'  or 'is considering their response' has no place in the verbal operant landscape. 

And... attempts to account for response variability in language have resulted in some rather fanciful behavior analytic formulations such as, 'Some intraverbals are not really intraverbals but are responses resulting from the additive effects of cascading stimuli generated by the question.' But what does this mean? How do questions generate cascading stimuli. I thought questions stimulated answers or other questions or comments. How does one go about verifying or falsifying such conjecture? What kind of experiments could be done to support such claims? Yet, such is the "science". This state of the art jibber jabber is a fabulous example of what happens when conceptual schemes are mixed; when explanations are given when no explanation is required except by the dictates of cultish, all inclusive, proclivities...everything can and must be explained by behavior analysis. 

Goals for language learning


> That children learn to use symbols (words, gesture) - according to the normative rules for their use. They learn the meaning of words such that they learn how to use them. Words express the concepts that sit under them. Thus, if a child uses and responds to the word "when" correctly we can say they have acquired the concept "when". Correspondingly, children need to learn that many words are used in different ways. Consider words like "bank", "ball", "band", "well", etc. Science can't account for these different uses any more than it can account for different uses of a knife. We learn by training in use. We learn technique. Whether something is a tact, mand or intraverbal is of little consequence in language learning.

> To learn the various language games within the practice.

-For example, to make 'moves' within a 'greeting game' allows for an indefinite number of appropriate moves. But, saying things like 'red shirt yellow hat', even when a greeter is wearing them (tact relations) just isn't a appropriate move in the game. But saying, "I hope you're feeling better", "How was your trip", "I've been waiting 25 minutes", are all appropriate moves when someone first says "Hi". No theory is needed to explain why these are reasonable responses. We've been socialized in greeting games and have learned what is appropriate in our cultural/linguistic practice. Of course saying that 'learning' has a place in mastery of practice is a perfectly acceptable thing to say. But saying so does not impute 'stimulus control', except if you are a behavior analyst.

> That children learn to do things with words for a purpose...that what is said has a point e.g., to direct, to redirect, to promise, refute, cajole, ask, invite, tell, warn, confirm, etc.

> Words are put into play within language games. Thus, the point of learning the names of things is not so that we can say that a child has "2500 tacts", but that words are put into play when games such as making a promise, negotiating, refuting, confirming, cajoling, inviting, greeting etc. Even if someone can name a ball... it is of little use unless the term is integrated within a practice of, for example, making a promise, i.e., 'I promise to give you the ball once I make a basket'. A name is of little use unless used for a purpose. Just like knowing the names of the pieces on a chessboard has little value unless one learns how to use those pieces; To learn techniques for their use. To learn a language, as Wittgenstein says, is to learn a technique.

>Teaching language to children with autism requires a constructional approach. We need to ask what is involved in being able to do "x" and what it takes to be able to do "x", what are the constituent abilities involved in "x" and how abilities relate. It's important to have in mind, to understand the reasons for teaching things; to consider how they will be used and what is necessary for engineering toward greater complexity and use. Rarely are things taught for their own sake. But unfortunately, in a VB paradigm, accumulating responses for their own sake is the goal, hundreds of thousands of them.


For example, what is the point of teaching children super-ordinate categories? What is the point of teaching children colors or other features of things. What is the point of teaching sorting and matching? What is the point of knowing the parts of things or regurgitating the functions of things in 'intraverbal training? One reason, not manifest in AVB, is the fact that these are abilities needed for constructing 'comparing' and 'contrasting' abilities which are involved in basic verbal reasoning abilities. In a verbal behavior paradigm, matching or sorting, answering arbitrary questions, learning FFCs and related intraverbals is the goal itself... not so that the information can be used for constructional purposes.

Let's  be clear


> Accumulating verbal units is not 'to do anything with words for any purpose'; Verbal units do not have a purpose. 'Purpose' has no place in a verbal behavior scheme as it smacks of 'mentalism'. In a behavioral world view, persons do not have agency but are at the mercy of the environment. Having verbal responses, within a verbal behavior conceptual system, is not to use symbols according to the rules for their use. Verbal operants exist independently of each other and have no relation to other units. Acquired verbal units are not used for constructional purposes but are the end goals themselves. Yet, children do learn these verbal units and such learning looks great on graphs.

> Verbal Behavior is an attempt to account for the behavior of the speaker in terms of operant relations /controlling relations (Skinner, 1957, p.2). It is not to account for or inform what is to be taught to persons hoping to learn a language/ to participate in a cultural/linguistic practice. Skinner was investigating behavior, not language. He makes this clear. 


> It must be pointed out and can't be overemphasized: The use of behavioral tools, procedures, strategies etc. need to be used for effective autism intervention and related language instruction. There is nothing more effective when employed properly. But, using these tools does not compel, nor should it require that practitioners subscribe to an AVB intervention approach or to behavioral philosophy (to which AVB adheres). 

> Mixing conceptual schemes results in utter confusion and nonsense. Adherence to AVB in efforts to remediate language deficits is misguided. If the example cited above about 'cascading stimuli' isn't enough to illustrate how confused and fantastical this all gets, consider another example. Carbon suggests that saying the word 'no' is to 'tact the absence of joint control'. Not only is this incoherent (it is not possible to tact something that is not there), but saying this adds nothing to how we should go about teaching children how the word "no" is used in our practice (to refuse, dis-confirm, reject etc.) Cultural practices are not reducible to science but are accounted for by cultural norms. Attempts to make language conform to science finds participants in that effort saying the most mysterious things.


Since verbal behavior is not about language, subscribing to a VB based intervention scheme necessarily denies children the benefits of language; being able to opine, reflect, ruminate, to think about the past or future, to perform logical operations that are linguistically mediated, to give reasons etc. And since such abilities smack of mentalism they are excised from considerations in a"Verbal behavior"intervention based scheme.


Adherence to AVB does not prepare children to participate in language practices, to learn the varied practice forms in language, to acquire concepts or to participate in the language games. These are not the intended outcomes of AVB, nor are they hoped for outcomes...amassing hundreds of thousands of independent verbal units across response classes is.

To say that someone understands the meaning of words does not require that we invoke 'minds'. In order to say that someone understands something, behavioral criterion are employed; we see that someone uses and responds to words correctly. Understanding is not in the mind-brain.


As language users, persons are doing things with words for a reason.  As agents, persons have the power to operate on the world...or not. The world is not arbitrarily acting on unwitting 'non-persons' to emit sounds. There are no hidden abilities or processes running in the background when we say that someone is doing something for a reason or even when the word 'mind' is invoked. The word 'mind' refers simply to an array of abilities.

> In order to ensure that language is addressed within an early intensive behavioral intervention framework, it's necessary to employ an ordinary language framework... the benefits of doing so are demonstrated in the groundbreaking work of Dr. Lovaas and his colleagues and others.(Lovaas, 1987, Eldevik, S., Hastings, R.P., Hughes, J.C., Jahr, E., Eikeseth, S., & Cross, S. 2010, Smith, T., Green, A., & Wynn, J. 2000, McEachin, J. J., Smith, T., & Lovaas, O. I. 1993, Howard, Jane S. , Sparkman, Coleen R., Cohen, Howard G., Green, Gina, & Stanislaw, Harold, 2005). Behavior analytic tools, procedures and strategies were employed in these efforts. These tools are unparalleled in their effectiveness for teaching. An AVB framework was not employed in these studies. There are no VB based studies demonstrating the superiority of VB over an ordinary language-based approach on a scale similar to the Lovaas study or similar studies. An ordinary language-based approach sets as its goals the things we do in language. It gets children in the game. Behavior analysts who insist that verbal behavior is language add nothing in efforts to advance language learning in youngsters struggling to learn a language but unverifiable and confused propositions, useless categorical systems and incoherent explanatory efforts that confound anyone trying to make sense of them (except for 'believers' who are able to suspend reason for the sake of 'the science').