Verbal behavior is a way of understanding language not by its structure (grammar, syntax, vocabulary) but by its function: why a person says what they say, and what effect it has on the people around them. The concept comes from B.F. Skinner’s 1957 book Verbal Behavior, where he defined it as behavior that is reinforced through the mediation of other people. In simpler terms, verbal behavior is any action (spoken, written, signed, or gestured) that works because another person responds to it in a learned way. This framework is now widely used in autism therapy and applied behavior analysis (ABA) to teach communication skills.
Why “Verbal Behavior” Instead of “Language”
Skinner deliberately avoided the word “language.” Traditional linguistics focuses on the structure of words and sentences: their grammar, meaning, and rules. Skinner wanted to do something different. He wanted to classify communication by what causes it and what it accomplishes. A child saying “water” because they’re thirsty is doing something fundamentally different from a child saying “water” because they see a lake, even though the word is identical. Linguistics treats those as the same word. Verbal behavior treats them as two completely different operants with different causes and different consequences.
This distinction matters practically. When a child struggles with communication, knowing they can say a word doesn’t tell you much. Knowing whether they can request things they want, label things they see, and answer questions tells you exactly where the gaps are and what to work on.
The Verbal Operants
Skinner broke verbal behavior into several categories called verbal operants. Each one is defined by what triggers it and what consequence maintains it.
Mand
A mand is a request. It’s driven by motivation: you want something, need something, or want something to stop, so you communicate to get that specific result. A child saying “cookie” because they’re hungry and then receiving a cookie is manding. The key feature is that the reinforcement matches the request. You ask for water and get water, not praise or a smile. Mands are the only verbal operant where the speaker’s internal state of wanting or needing is the primary driver, rather than something in the environment they can see or hear.
This is why many verbal behavior programs for children with autism prioritize teaching mands first. When a child learns that their words (or signs, or picture cards) reliably produce things they want, communication becomes immediately meaningful and rewarding.
Tact
A tact is a label. It’s triggered by something in the environment: an object, event, or property that the person can see, hear, smell, taste, or touch. A child who sees a dog and says “dog” is tacting. The reinforcement is typically social: someone acknowledges the label, says “yes, that’s a dog,” or simply responds. Unlike a mand, the child isn’t asking for the dog. They’re commenting on its presence.
Research on teaching tacts to children with autism has found that training works best when the nonverbal stimulus (the actual object or event) controls the response, rather than a verbal prompt like “What is this?” Children who learned to tact in response to objects alone, without a verbal question attached, showed better maintenance and generalization of the skill.
Intraverbal
An intraverbal is a verbal response to someone else’s verbal behavior, where the response doesn’t match the prompt. Answering “What’s your name?” with your name is an intraverbal. So is completing the phrase “red, white, and…” with “blue.” The response has no direct physical resemblance to the question; it’s controlled entirely by a learned history of certain words following other words.
Intraverbals are how conversations work. They’re also how people answer test questions, sing along to songs, and respond to social greetings. The complexity of intraverbal control increases with context. The word “red” alone could prompt dozens of responses: green, color, stop, fire. But the compound stimulus “red, white, and…” narrows the field dramatically in an American English speaker’s experience, making “blue” the dominant response. The more specific the verbal context, the more predictable the reply.
Echoic
An echoic is repeating what someone else says. A therapist says “ball” and the child says “ball.” The response matches the stimulus in both form and sequence. Echoic behavior is foundational for language learning because it allows a listener to practice and acquire new vocal responses by imitating a model.
Textual and Transcription
A textual operant is reading aloud: seeing the written word “cat” and saying “cat.” A transcription operant is the reverse direction, writing or typing what you hear or see in print. A student writing down a teacher’s spoken instructions is transcribing. Both involve a correspondence between written and spoken forms of language.
Autoclitics: Language That Modifies Language
Autoclitics are secondary verbal operants that modify the effect of primary ones. They’re the grammatical and qualifying elements that shape how a listener interprets what the speaker is saying. The word “rain” is a tact. But “I think it’s raining” adds an autoclitic (“I think”) that tells the listener about the speaker’s level of confidence. “I heard it’s raining” tells the listener the source of the information.
Skinner identified several subtypes. Qualifying autoclitics are driven by the speaker’s desire to influence the listener’s behavior in a particular way. Descriptive autoclitics are controlled by something about the speaker’s own behavior, essentially the speaker commenting on how confident or uncertain they are about what they’re saying. These small additions carry enormous weight in everyday communication, turning a flat statement into a nuanced one.
How Verbal Behavior Is Used in Autism Therapy
The verbal behavior approach, sometimes called Applied Verbal Behavior (AVB), is a framework within ABA therapy that organizes language teaching around the verbal operants rather than the traditional “receptive vs. expressive” categories. Instead of teaching a child to simply say words, therapists teach functional use of language across different operants. A child might learn the word “ball” as a mand (requesting it), a tact (labeling it), an intraverbal (answering “What do you play catch with?”), and an echoic (repeating it after a model). Each of these is treated as a separate skill that needs to be taught individually.
Practitioners typically start with mands because requesting gives the child immediate, tangible reinforcement for communicating. To build a mand repertoire, therapists arrange the child’s environment to create opportunities for requesting. They might hold a preferred toy in view but out of reach, withhold a missing piece of a puzzle, or pause a favorite activity mid-step. These moments of motivation create natural contexts where the child is prompted to communicate and then immediately receives what they asked for.
Once mands are established, therapists use transfer procedures to build other operants. If a child can already mand “cookie,” a therapist might present a cookie and prompt the child to tact it, gradually shifting the controlling variable from motivation to the visual stimulus. This systematic expansion across operants helps children develop flexible, functional language rather than rote responses tied to a single context.
Assessment Tools Based on Verbal Behavior
Two widely used assessments in verbal behavior programs are the VB-MAPP and the ABLLS-R. Both organize skill tracking around the verbal operants, but they differ in scope and target age range.
The VB-MAPP (Verbal Behavior Milestones Assessment and Placement Program), developed by Mark Sundberg, is a criterion-referenced tool designed for children with autism and other language delays. It divides development into three levels corresponding to 0 to 18 months, 18 to 30 months, and 30 to 48 months. Level 1 assesses early mands, tacts, listener skills, echoics, social behavior, motor imitation, and play. Level 2 adds intraverbals, classroom routines, and more complex listener responses. Level 3 introduces pre-academic skills in reading, math, and writing. The VB-MAPP also includes a transition assessment that helps determine what type of educational setting is appropriate for a child.
The ABLLS-R (Assessment of Basic Language and Learning Skills, Revised), developed by James Partington, covers 544 skills across 25 areas including language, social interaction, self-help, academics, and motor skills. It targets the developmental range of children up to about age six, roughly the skill level expected before kindergarten entry. It’s generally considered more comprehensive in scope, while the VB-MAPP is often preferred for focusing specifically on foundational language instruction.
Criticisms and Competing Frameworks
Despite its influence, Skinner’s framework has faced significant criticism. One major challenge comes from Relational Frame Theory (RFT), a newer behavioral account of language. RFT proponents argue that Skinner’s definition of verbal behavior is too broad: because it defines verbal behavior as behavior reinforced through another person, it doesn’t clearly distinguish verbal behavior from other social behavior. Asking someone to pass the salt and pushing someone out of the way both involve reinforcement mediated by another person, but only one feels like “language.”
RFT also criticizes Skinner’s definition of verbal stimuli. Skinner classified a stimulus as verbal if it was produced by verbal behavior, meaning the classification depends on the history of the person who produced the stimulus rather than the function of the stimulus for the person responding to it. RFT theorists see this as inconsistent with behavior analysis, which normally classifies stimuli by their effects on the individual organism.
RFT shares Skinner’s commitment to treating language as activity rather than a mental product, but it defines verbal events differently, focusing on the ability to derive relations between stimuli (for example, learning that A equals B and B equals C, then inferring that A equals C without direct training). This capacity for derived relational responding is, in RFT’s view, what makes human language fundamentally different from other learned behavior. While the debate continues, both frameworks remain active in research and clinical practice.