It’s been an ongoing debate on the ABA boards of the social media world: What the Function? (you see what I did there?)
“All behavior serves a function and it’s one of the Big Four: Attention, Escape/Avoidance, Access to tangibles, and Automatic.”
“No, no, there are 5 functions of behavior! Attention, Escape/Avoidance, Tangibles, Automatic, and ‘Control.’ It’s in quotes because it’s mentalistic and can’t be measured but somehow it’s also a function that we can measure.”
“Actually, no, you plebeians. There are 6 functions: Tangibles, Escape/avoidance, attention, communication, self-stimulation, and control.”
“To say there are only 4 reasons anyone does anything is absurd! We, humans, are so much more complex than that! That is a gross oversimplification of behavior!”
I see where all of that is coming from, but the argument is basically the same as when evolution deniers smugly ask, “if we come from monkeys, why are there still monkeys?” Saying it’s an oversimplification of behavior is itself an oversimplification of the functions. No, there really are 4. There just seems to be a whole lot of BCBAs coming out of grad school without a full understanding of them.
Warning: Nerdy Behavior Science Ahead.
This one’s just for us. Because despite my constant criticisms of ABA and the science of behavior, I still love it and this is a particular subject I’ve had some simmering nerd rage about for a while. Before we dive in too far, it’s important to establish that there’s definitely a possibility that I’m wrong and there really are more than 4 functions to behavior. That’s always a possibility, but I have yet to hear a compelling argument defending that idea. So here’s my take:
First, the functions themselves shouldn’t be seen as the only step. They are the first step. Insisting that they are not enough to cover behavior is not entirely inaccurate, but the claim is not usually meant this way. The functions themselves are huge umbrella terms with an even bigger variety of things underneath each of them. They’re not meant to be taken at face value, they’re meant to be a start of the full analysis.
Access and Control
The most poorly understood of them with the widest variety of interpretations would definitely be access. More commonly referred to as Access to Tangibles or simply Tangibles. It gets less accurate with each incarnation. The main focus should not be the tangibles, it should be the access. Tangibles are most often thought of as items or activities, but what about people? Sometimes the reinforcing consequence has nothing to do with objects and everything to do with the one providing them.
Now you might be thinking: Isn’t that attention? Well, not exactly. It depends on what purpose that person serves. Again, we are narrowing our scope too far. We think of people as being vehicles of attention, but is anyone really that extroverted? Haven’t you ever just hung out with someone? No one is talking, you may not even be doing the same thing. That person is just in the room with you and that’s enough. No attention is occuring. That person is simply present in the vicinity.
Let’s talk about vicinities. What about preferred rooms? Again, not to necessarily get anything from said room, just to be in that room. You’re accessing an array of tangibles and intangibles collectively referred to as a room. Maybe it’s got particular lighting that is more comfortable, or something about the location or walls is better at blocking out sound, making that room quieter than other rooms. These are especially important considerations when dealing with neurodivergent folks who may have sensory sensitivities.
Of course, that leads to a whole slew of other conversations: A negative vs. positive reinforcement conversation. Is it access to the room itself or the escape from aversive stimuli it provides? Sometimes it’s impossible to know, so we just say it’s multiply determined or just pick one.
Then there’s “control,” complete with quotes (or air quotes if you’re using mouth words). The first time I had this concept explained to me, I was roughly a year out of grad school, and it broke my brain. Not in the way the person explaining it to me might have expected, though. I’m listening to them struggle to explain the whole thing and struggling to keep up with their perspective. Then, after politely letting them finish, I asked: “wouldn’t that just be access to choices?”
And then their brain broke in the way they had previously expected mine to.
What does the “control” function look like? It’s the person who insists on being the one to initiate the actions. You can’t close the door, they have to. If you close it, they will open it again and then close it.
They are the ones who refuse to do exactly what you say.
You say “sit in the blue chair,” so they sit in the red one.
How do you address disruptive behavior that is maintained by control? Well, there are a few options with varying levels of shitty and not. Sometimes it involves blocking, redirecting, or physically prompting the person to do exactly as you said and only as you said. Or, it could be forcing them to accept the environmental setup as it is. Maybe they can even “earn” the opportunity to close the door or sit in the red chair if you’re very clever and good at controlling reinforcers.
If you’re not sure what’s wrong with that, welcome to the blog. Feel free to look around.
It can also be prevented by offering choices: Do you want to sit in the red chair or the blue chair? You can teach less disruptive or stressful (on the learner) ways of expressing the desire to control, such as teaching the person to say “I want to sit in the red chair,” whatever that looks like for them. These are effective because the person is getting access to the opportunity to make a choice or to in some other way influence– “control” if you will– the environment.
Medical, Sensory, and Automatic
Honestly, I don’t understand the reason for breaking these up. They’re all automatic. The misunderstanding of the functions as oversimplified may lead one to feel like all of these need to be specified, but it doesn’t make them all separate functions. If a function is medical, the behavior is the result of a medical condition, usually escape or avoidance of associated discomfort. Sensory can also be a result of a medical condition, but not necessarily.
Either way, the behavior is being maintained by the very act of engaging in the behavior to begin with. That is the definition of automatic reinforcement. Breaking these up as their own functions just seems like a failure to fully understand the meaning of the function categories. The behavior is maintained by automatic reinforcement, but that doesn’t just mean stimming.
Either way, you are always ruling out medical causes, right?
We are not suppressing automatically reinforced behavior that isn’t harmful, right?
Communication
There is a popular idea that all behavior is communication. I don’t agree. Not just for the fact that it doesn’t fit with the scientific definition of behavior, but because sometimes that simply isn’t the case, not even for behavior that is challenging to caregivers.
Don’t get me wrong: It is often a component, and teaching communication is never a bad thing, so long as you can be sure the communication you are trying to teach is truly functionally equivalent. Teaching someone to say “hey, I need attention,” is not going to be functionally equivalent if the kind of attention reinforcing the behavior is negative. The real question in that case is why is that type of attention — scolding, yelling, etc.– reinforcing? But that’s a different conversation.
Communication itself is also made up of a variety of complex behaviors, including verbal and nonverbal, and vocal and non-vocal. The functions with these can also vary, especially when you get into autoclitics, in which mands may be embedded into one’s verbal behavior in order to indicate a change in function: A particular type of exchange or interpretation on the part of the listener (access again). Communication cannot itself be a function because communication also serves different functions.
The Benefit of Extra Descriptors
Additions to the functions are not without merit. Because the functions are such large categories, it’s important to understand all of the different components in order to best describe the actual function of the behavior you’re analyzing. The argument is simply that each of these components is not itself a larger function. They are still a part of the 4 main ones. Think of it like categorizing life: A bear and a bee are extremely different, but they are both part of the animal kingdom.
Scolding, praise, eye contact, simply changing one’s behavior, talking about a behavior to someone else, etc. are all components of attention, but the behavior would still be maintained by attention.
People, locations, and options are still access. Access as a function may not be for a specific tangible, though tangibles may still be involved. The ability to break down tangibles into all of its possible components is also going to be helpful in fully describing the reinforcing consequences. Similarly, a lack of understanding for the complexity and forms possible under the “access” function can also result in inadequate measurement, which can lead to those horrible functional analysis graphs that look like a failed roller coaster project. When we cannot properly define what we are measuring, we cannot properly analyze, and by proxy treat, the behavior.
Assigning mere attributes as functions themselves leaves less room for further understanding the conditions under which the behavior is more or less likely to occur. Too many functions encourages us to stop digging into the specifics of what’s maintaining that behavior, whereas understanding their breadth encourages continued analysis all the way to the end. You put a behavior plan in place with functions that are both too specific and not specific enough, and you end up crying in front of a zig-zaggy graph with no idea why your behavior plan isn’t working.
These are the things that keep us up at night.