Identify the Role of Multiple Control in Verbal Behavior
You’re in session with a learner who says “cookie” perfectly when you ask, “What do you want?”—but walks right past cookies on the shelf without naming them. Or a child requests a break only when frustrated in the presence of a specific adult, but nowhere else. These inconsistencies aren’t usually signs of poor learning. Most often, they’re signs that multiple control is at work—and once you see it, your assessments and interventions become far more precise.
Multiple control in verbal behavior means that a single verbal response is usually governed by more than one antecedent variable. Rather than a word or phrase being “controlled” by one source, it emerges from the combined influence of a motivating operation (MO), a discriminative stimulus (SD), a prompt, or some combination of these. Understanding how multiple variables converge to shape what a learner says—or how a single antecedent can evoke many different responses—changes how you assess communication and plan intervention. Without this lens, you risk designing treatments that work beautifully in your clinic but fail to generalize. With it, you build durable, flexible language skills that actually transfer to real life.
What Is Multiple Control in Verbal Behavior?
In Skinner’s framework, verbal behavior is almost never controlled by a single factor alone. Think about why you say something in any given moment: your words emerge from the interplay of what you see (an SD), what you need or want (an MO), what someone has just asked you (another SD), and the reinforcement history behind your words. This is multiple control—and it’s the norm, not the exception.
Convergent multiple control describes what happens when several antecedent variables come together to evoke one response. Imagine a child who is hungry (MO) and sees a cookie on the table (nonverbal SD). A therapist then asks, “What do you want?” (verbal SD). The child says “cookie.” That single word is under the joint influence of the child’s hunger, the sight of the cookie, and the question. Remove any one of those variables—say the child is no longer hungry, or you ask a different question—and the response becomes less likely.
Divergent multiple control is the flip side: one antecedent controls multiple different responses. You show a learner a picture of a dog, and they might say “dog,” “animal,” “puppy,” or “bark” on different occasions. The same stimulus evokes different topographies depending on prior reinforcement, context, or what the learner has been primed to attend to. This diversity is actually healthy—it’s part of what makes language flexible and adaptive.
Beyond these two primary types, related concepts add nuance. Thematic control occurs when a cue and response share a common meaning or topic rather than a direct physical resemblance. Formal control involves responses that resemble their stimulus (like echoing words you hear). Supplementary stimulation refers to prompts or hints that add strength to a response already in the learner’s repertoire. And joint control—a specific case of convergent control—involves two stimuli that must co-occur to fully evoke a response, often seen in complex language tasks.
The key is recognizing that multiple control involves functional relations, not just temporal coincidence. If a response happens to occur when two things are present, that doesn’t mean both are actually controlling it. True multiple control means you can systematically change one variable and see the probability, latency, or form of the response shift in predictable ways.
Why Multiple Control Matters in Assessment and Intervention
Here’s the practical reality: understanding multiple control prevents brittle skills. A learner who produces a label only when prompted, or only in the presence of a specific person, hasn’t really learned the skill—they’ve learned a narrow, context-bound response. When you recognize that multiple controls are at play, you can design assessments and teaching that address all of them.
In functional analysis, this means more than observing a behavior in one or two conditions. You need to test the suspected antecedents both separately and together. Does the response occur without the prompt? Without the motivating operation? Without the adult asking the question? Manipulating these variables reveals which ones are truly necessary. Once you know what’s controlling the response, you can plan fading that actually leads to generalization—rather than fade away the prompt and watch the response collapse because the underlying MO or SD was never strong enough.
For intervention, recognizing multiple control changes your teaching approach entirely. Instead of focusing on a single operant or function, you plan around all identified controls. If a child requests “break” only when they’re frustrated and when prompted, you don’t just practice the mand in isolation. You build a functionally equivalent communication repertoire where the child can request break across varying levels of frustration, in multiple contexts, and without adult scaffolding. You also attend to whether the learner needs training on related skills—perhaps a tact for what’s causing the frustration, or an intraverbal response that gets at the same function.
Understanding multiple control also makes you more respectful of the learner’s communicative intent. When you see the full picture of what’s controlling a response—not just the prompt you’ve provided, but the natural MO, the relevant SD, the learner’s history—you can design interventions that honor that intent rather than override it.
Types of Multiple Control: Convergent and Divergent
The distinction between convergent and divergent multiple control is foundational, and it shapes how you think about almost every verbal behavior you see in session.
Convergent multiple control means multiple variables are pointing toward the same response. A child sees a preferred toy, wants it (MO), and hears you ask “What do you want?” The word “toy” emerges from all three sources at once. This is incredibly common—most functional communication blends a mand (request driven by an MO) with a tact (label driven by the presence of the object). In real life, language isn’t usually “pure.” Recognizing this convergence helps you avoid the trap of thinking that a single antecedent is enough.
Divergent multiple control describes a single antecedent leading to a variety of possible responses. Show a learner a dog, and over time, they might produce “dog,” “animal,” “puppy,” “furry,” “tail,” or “bark.” None of these is “wrong”—they’re all related to the stimulus, and which one emerges depends on the learner’s reinforcement history in similar contexts. Divergent control supports richer, more flexible language. It’s why a skillful communicator can say the same thing in many ways, and why learners whose divergent control is weak often sound scripted or repetitive.
In practice, you’ll find both types operating simultaneously in a single learner. The key is recognizing which is which, so you can design interventions accordingly.
Key Features and Defining Characteristics
Multiple control has several features that distinguish it from other forms of stimulus control or prompting effects:
Multiple functional variables combine to evoke a single response. This isn’t about physical features lumped together—it’s about the independent functional contribution of each variable. An MO works differently than an SD; a prompt works differently than a naturally occurring cue. When multiple control is present, each can be manipulated separately, and each change alters the response in measurable ways.
The effects can be additive, hierarchical, or conditional. Sometimes a response requires all variables to be present (hierarchical). Sometimes any combination of two will do (additive). Sometimes one variable only matters in the presence of another (conditional). Your job is to figure out which pattern is operating.
The response depends on demonstrable functional relations, not just timing. A response that happens to occur when two things are present doesn’t prove both are controlling it. True multiple control requires that you can manipulate each antecedent and observe predictable changes in the response.
Observable indicators include changes in probability, latency, or form when one antecedent is altered. If you remove the MO and the response disappears even though the SD is still present, the MO was a necessary component. If you remove a specific prompt and the response stays strong, that prompt wasn’t actually controlling it.
When and How to Use Multiple Control in Practice
You reach for multiple control analysis at specific decision points in your work.
When a verbal response is inconsistent across contexts, multiple control is the first place to look. The child labels objects perfectly with their mom but not with their dad. They request a favorite item at home but not at school. Before assuming the learner “hasn’t learned it,” ask: what’s different across these contexts? Which motivating operations are present or absent? Which discriminative stimuli differ? Often, the inconsistency reveals that the response is under convergent control—and the missing variable is why it doesn’t generalize.
When prompting or fading fails, multiple control is often the culprit. You fade the prompt, the response collapses, and you’re confused because the learner “knew” the skill. The issue: the learner knew the skill in the presence of that prompt. The underlying MO or SD was never strong enough to maintain the response on its own. Recognizing this, you shift strategy. Instead of just fading the prompt, you strengthen the natural controls—arranging for the MO to be more salient, creating opportunities for natural reinforcement, or teaching in a context where the SD is unmissable.
When planning multi-component communication, understanding multiple control helps you build richer teaching. If you want a learner to request an item (mand) and label it (tact), and you want both to occur smoothly, you’re working with convergent control. You might work in a context where the child both wants something and has a reason to label it—creating a natural blend that honors both functions.
A practical scenario: a learner names colors correctly when you point to them and ask, “What color is this?” but doesn’t spontaneously name colors in the environment. Multiple control suggests the question (SD) is necessary. Your next move isn’t to shame the learner for not naming unprompted. It’s to gradually fade the question while building in natural consequences for color naming, or to create a communicative context where naming colors matters to the learner.
Real Examples in ABA Clinical Practice
Example 1: The Mand-Tact Blend
A child is hungry (establishing operation for food), sees a cookie on the counter (visual SD), and a therapist asks, “What do you want?” (verbal SD). The child says “cookie.”
What’s controlling this response? All three variables. The hunger makes the cookie reinforcing. The sight of the cookie provides a nonverbal cue. The question provides a verbal cue and clarifies the response topography needed. Remove any one—say the child just ate, or the cookie is hidden, or you ask “Where is it?”—and the response becomes less likely or changes form.
This is convergent multiple control in action. The learner is both requesting (controlled by MO) and labeling (controlled by the object itself). Recognizing this, you’d plan intervention to strengthen each component—building the mand even without the object present, and the tact even when the child isn’t hungry.
Example 2: Converging Controls with an Autoclitic
A learner is hungry, sees a picture during naming time, and the therapist asks, “What’s this?” The learner says “apple please.”
Here, multiple controls converge on a single utterance. The picture is a tact SD. The hunger is an MO that’s reinforcing food words. The question is a verbal SD. But the learner has also added “please”—a modifier that softens the request and functions as an autoclitic. The autoclitic itself is under control: the learner’s reinforcement history with politeness, the therapist’s prior approval of “please,” and the social context all shape whether it appears. The learner isn’t just naming; they’re communicating with nuance.
Common Mistakes and Misconceptions
A frequent error is assuming a single antecedent controls a verbal response when multiple variables are actually present. A learner says a word correctly under three conditions: when you use a specific prompt, when they want something, and when the object is visible. It’s tempting to credit one variable—often the prompt, because that’s what you intentionally introduced. But the response may depend on all three. Testing them separately reveals the truth.
Another trap is confusing temporal sequence (chaining) with simultaneous functional control. If a learner hears a prompt, then sees an object, then responds, it looks sequential. But that’s not the same as true multiple control. In genuine convergent control, the variables operate together at the moment of response, not in a chain. The distinction matters because your intervention differs. If control is chained, you fade the chain. If control is convergent, you strengthen each source independently.
Labeling any multiword utterance as “multiple control” without showing separate functional influence is another common slip. A learner says “I want cookie.” That’s two words or more, but it might be controlled by a single operant—a complex mand, where “I want” is part of a learned form, not evidence of two separate controlling variables. Multiple control requires demonstrating that distinct variables each contribute.
Finally, confusing multiple control with joint control, or conflating compound stimulus control with general multiple control, muddles the picture. Joint control is a specific mechanism where two stimuli must co-occur to fully evoke a response. A compound stimulus is a combination of physical features that has acquired control as a unified unit. Multiple control is broader: it includes joint control as a special case, but also any situation where separate functional variables contribute to a response.
Ethical Considerations and Best Practices
Working with multiple control, especially when it involves manipulating motivating operations, carries ethical weight. Any deliberate arrangement of deprivation or satiation—even mild ones—must be approached with careful consideration of the learner’s welfare, informed consent from guardians, exploration of less restrictive alternatives, and attention to whether the approach is socially valid.
If you’re planning an intervention that involves managing MOs—say, arranging mild hunger to strengthen motivation to request food, or timing lessons when a learner is likely to want a break—that learner’s guardian must understand what you’re doing and why. Transparency and consent matter.
Autonomy and dignity also matter. When you recognize that a learner’s response is under convergent control, you’re in a position to understand their communicative intent. Design interventions that honor that intent, rather than training narrow scripts that remove natural sources of control. If a child naturally wants to request a toy, teach requesting for that toy, not something you think they “should” want.
Document your reasoning about multiple control clearly. If you suspect convergent or divergent control, note which variables you tested, how you tested them, what the results were, and what you concluded. If multiple control analysis changes your intervention plan, record that. Seek supervision, especially if the suspected control involves MO manipulation or if your testing was limited.
Practice Questions to Solidify Your Understanding
Question 1: A child says “cookie” only when seeing a cookie AND when the caregiver asks, “What do you want?” The response doesn’t occur if the cookie is hidden, even with the question. Which best describes the control?
Answer: Convergent multiple control. Both the EO (hunger/deprivation for the food) and the verbal SD (the question) are necessary. Remove either, and the response becomes less likely.
Question 2: A single picture of a dog evokes several different responses from a learner: “dog,” “animal,” “pet,” and “furry.” Which term fits?
Answer: Divergent multiple control. One antecedent (the picture) is controlling multiple different responses. This reflects the learner’s rich reinforcement history and flexible language repertoire.
Question 3: During assessment, you remove an MO and the verbal response disappears, even though the SD is present. What does this suggest?
Answer: The MO was a necessary component of convergent control. Your intervention should strengthen the MO-SD relationship or build in alternatives that don’t require that specific MO.
Question 4: A learner uses a grammatically modified utterance like adding “please” to a request. How should you classify the added phrase?
Answer: An autoclitic that modifies the primary verbal response and its controlling variables. Recognizing this, you understand that the learner is doing more than just requesting; they’re negotiating the social context of the request.
Related Concepts That Deepen Your Understanding
Mands (requests) are often influenced by MOs and can co-occur with tacts and other operants, creating convergent control. Tacts (labels) are controlled by nonverbal SDs but can blend with mands in real communication. Intraverbals (verbal-to-verbal responses) can interact with MOs and SDs, sometimes creating unexpected multiple controls. Autoclitics are secondary verbal operants that modify primary responses; they’re controlled by the speaker’s certainty, emotional state, and history, adding another layer of control architecture.
Motivating operations (MOs) alter the value of consequences and are fundamental to manding. Discriminative stimuli (SDs) set the occasion for reinforced behavior and are fundamental to tacting and many other operants. When MOs and SDs converge, you have convergent control. Prompting and prompt fading must account for multiple sources of control; fading a prompt before the natural SD or MO is strong enough leads to brittle, prompt-dependent skills.
Summary and Next Steps
Multiple control is at the heart of why language is both rich and context-dependent. Most verbal responses aren’t controlled by one thing—they emerge from the interaction of motivating operations, discriminative stimuli, prompts, prior reinforcement, and autoclitic modification. Convergent control (multiple variables → one response) and divergent control (one variable → multiple responses) are the two primary patterns.
When you see inconsistent responding, prompt dependence, or brittle skills, multiple control is often the explanation. Rather than assume the learner hasn’t learned, ask: what’s different across contexts? What antecedents are present in some situations and absent in others? By systematically testing variables and documenting the results, you build a clearer picture of what’s actually controlling the response. Then, your intervention can address all the functional sources—not just the prompt you introduced, but the natural MO, the relevant SD, and the broader communicative context.
This approach is more respectful of the learner’s autonomy and communicative intent. It’s also more effective. Skills built on a foundation of multiple, naturally occurring controls generalize better, last longer, and feel less like obedience and more like genuine communication.
As you move forward, reflect on your current learners. Are there responses that seem inconsistent across contexts? Where might multiple control be operating? How might a deeper analysis of the controlling variables change your next step in teaching?
