Manding for Information: How to Teach a Child to Ask 'Why?'

Manding for information matters because a learner who can ask "why" stops waiting on the world to explain itself, and Valentino and Padover (2019) showed it is also one of the trickiest WH-mands to set up cleanly. The classic example is a therapist walking in with a silly top hat on, hoping the child looks up and asks why. That works once. The problem is the kid who only cares about dinosaurs. If the top hat is not a Tyrannosaurus, he does not care, and you have just burned a trial on a question he never wanted to ask.

That is the whole problem with teaching "why" in mand training. The EO (the establishing operation, the reason a person wants something at this moment) has to be real curiosity, not adult-scripted curiosity. And the second you answer the question, the curiosity is gone.

Why "why" is the hardest WH-mand to teach#

Most WH-mands give you a fresh setup every time. "What is that?" can target a new object. "Where is it?" can hide a different item. "Who has it?" can change the holder. With "why," the answer itself ends the trial.

once you provide the information, once you satisfy that question mark, all of a sudden there's no longer an EO. In fact, it would be called an EO absent condition because there already has been an answer to that question. From the talk — Matt Harrington

That is why a single contrived scene burns out fast. Once the child knows why you are wearing a top hat, the top hat is no longer interesting. If you reuse the same setup tomorrow, you are not teaching a mand any more. You are teaching a script.

What an EO looks like for a question (curiosity, not hunger)#

In basic mand training the EO is easy to picture. A kid is thirsty, so juice is the reinforcer. A kid wants a break, so escape from the table is the reinforcer.

For "why," the reinforcer is information. The EO is a small spike of curiosity in the moment. Something does not fit. Something looks off. Something is unexpected. Until that mismatch shows up, the learner has no reason to ask anything.

That is the work you have to do before you ever run a trial. You are not just placing a stimulus. You are setting up a contrast between what the child expects to see and what they actually see. Top hats only work as a contrast if the child noticed that adults do not usually wear top hats.

The silly top-hat method from Valentino and Padover (2019)#

The Valentino and Padover (2019) procedure puts the EO right out in the open. A familiar adult walks in wearing something obviously out of place. A wig. A baseball mitt on one hand. A traffic cone on their head. The child looks at the adult, the adult does not explain, and the child gets prompted to ask, "Why are you wearing a top hat?"

That is the bones of it.

they were teaching the man why, and they had the therapist put a silly thing on their head, right? So he would put a top hat on the head, and the child would ask, why are you wearing a top hat? From the talk — Matt Harrington

The reinforcer is the answer. "Because I won a costume contest." "Because my dog ate my real one." The answer has to feel real and a little surprising, because that is the consequence the mand is recruiting. A dull "I just felt like it" trains a flat mand.

The procedure also assumes the kid can already mand for simpler things and can echo a short prompt. If both of those are shaky, fix them before you touch "why."

Why you need 15 to 30 prepared scenarios, not 2 or 3#

The number that catches people off guard is how many setups you actually need.

before you set up and you decide to run trials on teaching why, you should have in your back pocket 15, 20, maybe even 30 different why scenarios that you're going to run through so that you can always have an environment where an establishing operation is present. From the talk — Matt Harrington

Two or three is the most common mistake. A team picks the hat, the wig, and the rubber chicken, runs them every session for a week, and by Thursday the child is reciting "Why are you wearing a top hat?" without looking up. That is not a mand. That is a tact of "Wednesday afternoon."

A useful scenario bank covers different categories. Clothing that does not belong (a winter glove on one hand). Objects in wrong places (a plastic banana inside a backpack). Sounds (the therapist humming a movie theme on a loop). Actions (drawing with the wrong end of a marker). Smells, if the kid is sensitive to those. Items that show up at strange times, like a flashlight at noon.

Plan for fifteen on day one. Add a few each week. Retire any that the child has answered three times in a row, because the EO is gone.

Using the child's special interest to load real curiosity#

This is the piece that most BCBAs need to hear, and it is where many "why" programs quietly die.

A generic top hat is a contrast for most kids. For a kid whose whole inner life is dinosaurs, a top hat is noise. He has already filtered it out.

somebody, it was somebody who wasn't even involved in ABA. I was just venting a little bit. And they were like, does he care? We ended up changing all of our scenarios to relate back to dinosaurs, which was this individual special interest. And everything totally changed. From the talk — Matt Harrington

So the new top hat became a stuffed T-Rex on the therapist's shoulder. The wrong-place object became a brachiosaurus inside the snack bin. The strange sound became the therapist quietly singing the Jurassic Park theme while filling out the data sheet. Each one snapped the kid's head up. Each one earned a real "Why?"

When you build a scenario bank, write a column next to each setup labeled "fits the kid." Trains. Police cars. Specific YouTubers. Specific songs. If the contrast does not connect to that column, swap it out. You are not designing for "a child." You are designing for one child whose curiosity has a known shape.

Checking that the mand is discriminated, not just rote#

A mand has to be under the right control. For "why," that means the question only shows up when something unexpected is happening, and it does not show up when nothing unusual is going on.

A quick check. After a few sessions of "why" trials, run a probe with no setup at all. Walk in normal clothes, sit down, do regular table work. If the child looks up and asks "Why are you wearing a top hat?" anyway, the mand is not discriminated. The child has learned that asking the question makes adults talk. You have a rote string, not a mand for information.

Two more probes worth running. First, swap in a new staff member with a fresh, never-used contrast. If the child asks "why" in response to that new setup, the mand is generalizing. Second, run a "no contrast" trial in a different room. If the child stays quiet, the discrimination is holding.

If the mand fails any of these probes, the fix is usually scenario variety plus more EO contrast, not more prompting.

When to move from why to who, what, when, and where#

Once "why" is stable across at least ten unique scenarios, two settings, and two adults, you can layer in the other WH-mands. Order matters less than reinforcer design.

"What" is often the easiest next step because the answer (a name) does not destroy a long-term EO. The kid asking "What is that?" about a strange object still wants the object after they hear the name.

"Where" needs a hidden item the child wants to retrieve, so the reinforcer is finding it.

"Who" needs an unfamiliar voice, footstep, or hand entering the scene, so the reinforcer is identifying the person.

"When" is the hardest after "why," because waiting for a future event is hard to reinforce in the moment.

Mix them in slowly. Keep "why" trials in rotation so the skill does not regress while the new questions take hold.

Frequently asked questions#

What if my client asks "why" but doesn't seem to listen to the answer? That usually means the answer is not the reinforcer. The mand is being maintained by attention or the act of asking, not by information. Check whether your answers are short, surprising, and tied to the child's interest. A flat "because I wanted to" trains a flat mand. A two-second pause and a real reason ("because my brother dared me") gives the information value.

How is teaching "why" different from teaching "what is that?" "What is that?" can repeat across hundreds of unknown items because each new object is its own EO. "Why" cannot, because the answer kills the curiosity for that specific scene. That is why "what" runs on a long item list while "why" runs on a deep scenario bank.

How many trials per scenario before I move on? A rough rule is three correct, independent mands across two sessions, then retire the scenario for at least a week. If you keep running it, the question becomes rote. If you cycle it out, you can bring it back later as a probe to confirm the skill is still under EO control.

Keep going#

Manding for information is one of the most powerful skills in a verbal behavior program, and "why" is the gate. Watch the full session for the prerequisite checks, the prompt-delay schedule Matt uses, and the way he layers in the other WH-mands without losing what the child already learned.