
Last updated 2026-07-10
TL;DR
Kids usually answer simple 'what' questions by age 2 to 3 and harder ones by 4 to 5. You teach it by starting with objects the child can see, using errorless learning to build confidence, offering forced choices, and practicing daily inside normal routines. Late talkers and autistic kids often need the skill broken down and taught directly, with visual supports and sometimes AAC.
Most children start responding to very basic 'what' questions somewhere around 18 to 24 months, when the question is just "What's that?" with a finger pointed at a dog and the expected answer is one word: "dog." Not sophisticated, but a real exchange. By age 3, a typically developing child answers questions about objects, body parts, and familiar actions without much prompting, and by 4 to 5 they can handle something more abstract, like "What do you do when you're hungry?" or "What happened at school today?" [1]
If your child is past age 3 and still isn't reliably answering simple questions about things right in front of them, mention it to a speech-language pathologist. It doesn't point to anything dire; it usually just means the skill needs direct, intentional teaching. The American Speech-Language-Hearing Association recommends an evaluation whenever a caregiver has a concern about language development, and age 3 is a common point at which families get referred. [2] Late talkers, autistic children, and kids with apraxia of speech often hit this milestone later, or need a different teaching route altogether, which is what the rest of this article covers.
Why answering 'what' questions is harder than it looks
Parents assume that if a child understands language, answering questions should come naturally. It doesn't. Answering is a separate task from understanding, and it can break down at several points. To answer "What is that?", a child has to recognize the question as a request for information rather than a comment, pull the right word from memory, say it at the right moment, and hold back other responses. Four steps, and any one of them can stall.
For autistic children, questions can also be socially confusing. Research on autism and language shows many autistic kids respond to 'what' questions with echolalia, repeating part or all of the question back, or with silence, not because they don't know the answer but because the social demand of being asked something is unclear or uncomfortable. [3] If that's the pattern you're seeing, read about echolalia first, since the approach is different from closing a plain vocabulary gap.
Children with word-retrieval difficulties, including those with childhood apraxia of speech, may know the answer perfectly well but struggle to get the motor plan for the word out in real time; pushing them to hurry or correcting them repeatedly tends to make retrieval harder, not easier. Most often, though, the real issue is simpler: nobody has broken the skill down and taught it yet. That's also the easiest one to fix.
Which 'what' questions come first
Not all 'what' questions are equally hard, and teaching them out of order is one of the most common mistakes parents make at home. Roughly, the progression runs from easiest to hardest:
| Level | Example question | What's required |
|---|---|---|
| 1. Object naming | 'What is that?' (pointing at visible item) | Single word, item is present |
| 2. Category / function | 'What do you eat?' or 'What goes on your feet?' | Word retrieval, basic concepts |
| 3. Action | 'What is he doing?' (looking at a picture) | Verb retrieval, ~ing forms |
| 4. Attribute | 'What color is this?' or 'What shape?' | Descriptive vocabulary |
| 5. Experience / event | 'What did you do today?' | Episodic memory, sentence structure |
| 6. Hypothetical | 'What would you do if...?' | Inference, abstract thinking |
Nearly every child should start at Level 1, even one who can already label objects independently, because it produces quick wins and builds the question-then-answer rhythm before the vocabulary demands climb. [4] Most parents, understandably, start at Level 5 ("what did you do today?") because that's what they actually want to know, but it's also where nearly every child with a language delay gets stuck. Set those questions aside for later.
Teaching the skill step by step
Here's the sequence SLPs use in practice, simplified for home, and it works across ability levels, including for children who are minimally verbal.
Start by teaching the format before the content: before worrying about whether your child knows the answer, show them what a 'what' question sounds like and that it calls for a response. Model it yourself, out loud, with no pressure on the child to reply: "What is this? It's a cup." Do this dozens of times across the day. This kind of self-talk modeling is supported by research on naturalistic language intervention. [5]
Next comes errorless learning for the first answers, meaning you set the question up so the child can only land on the correct response. Hold up a ball, say "What is this? It's a..." and pause with expectant body language. If they don't finish it, say "ball" yourself, wait, then try again. You're building the retrieval pathway without attaching any failure to it. That's not handing them the answer; it's teaching.
Once errorless trials bring some wins, move to forced choices: "What is this, a ball or a book?" Two named options, one correct, which cuts the retrieval demand while still making the child process the question and choose. Fade to one choice over time, then to no choice at all.
Use real objects and visuals wherever you can, since real objects work better than pictures and pictures beat nothing. If you're working on food words, do it in the kitchen with actual food in view; visual context supports word retrieval far more than asking in the abstract. [6]
Once a child answers reliably in structured practice (say, correct about 80% of the time across three days), carry the same question types into daily life: bath time, meals, playing outside. Natural routines give repetition without turning into a drill. Then start the cycle again with new vocabulary, animals, then clothes, then actions, then attributes, and expect to lean on errorless learning again each time a new category starts. That's normal, not a setback.
Visual supports worth using
Visual supports work because they take the load off working memory: instead of pulling a word entirely from inside their head, a child can look at a picture, symbol, or written word that triggers the retrieval. This helps autistic children and kids with significant language delays the most, and none of the options below require special training.
A choice board, a card or tablet screen with 2 to 4 pictures, lets a child point instead of speak. Ask "What do we need to wash hands?" and the board shows soap, a towel, a cup, and a fork; pointing to soap counts as a real answer and builds the habit of responding with correct meaning. First-then boards don't help much with 'what' questions directly, but they lower anxiety around structured tasks, which helps a child stay engaged long enough to participate.
For children who are minimally verbal or who use augmentative communication, answering a 'what' question often means selecting a symbol on a device, and that response deserves exactly the same reinforcement a spoken answer would get. [7] You can read more about how AAC devices support language learning at every stage. A small photo album of your child's own life (their dog, their bedroom, their favorite toy) is also a strong object-naming tool: personal photos pull more engagement than generic pictures and cut down on guessing.
How many repetitions it actually takes
Nobody has solid population-level data specifically on 'what' question answering. The closest research comes from behavioral intervention studies, which generally find children with autism need somewhere between 20 and 200 massed practice trials to reach criterion on a discrete skill, depending on the child and how complex the target is. [8] That's a wide range, and it's the honest picture.
What it means in practice: short daily sessions of 5 to 10 minutes beat one long weekly session. Three minutes at breakfast, three at bath time, three at bedtime adds up faster than it sounds. The variable that matters most is reinforcement, whatever the child finds motivating (praise, a small preferred toy, a bite of a favorite food) should follow a correct or approximated answer right away; reinforcement delivered a minute later teaches almost nothing, while reinforcement within a second or two builds the association.
Parents sometimes worry they're bribing their child to talk. They're not, they're teaching. The treats and praise fade out naturally as the behavior becomes habit and the social payoff of communicating turns into its own reward.
What if my child answers with echolalia instead of the real answer?
If you ask "What is this?" and your child says "What is this?" right back, that's echolalia, and it calls for a different response than simply repeating yourself.
Don't tell them they're wrong, and don't just ask again the same way. A few things work better. Switching to a fill-in format ("This is a...") with a clear pause often helps, since it's easier than answering a question outright for kids who echo: it strips out the question syntax that may be triggering the repeat. Cutting down your language load helps too. Shorter, simpler prompts leave less to echo, so instead of "Can you tell me what this is?" try just "What is it?" And give it time: some children need five to ten seconds to process and produce an original response, and the echolalia is filling that silence while they work. Wait it out before piling on more prompts.
Don't try to erase the echolalia at the same time you're teaching answering. Those are two separate goals, and an SLP can help you prioritize between them. Echolalia often carries communicative function even when it looks like nonsense, and cutting it without replacing it with functional communication tends to backfire. [3] Our closer look at echolalia meaning goes into how to read different patterns.
Does this belong in speech therapy, or can I do it at home?
Home practice matters enormously. Research on naturalistic language intervention consistently shows that parent-implemented strategies between sessions improve outcomes compared to therapy alone. [9] A child who sees an SLP for 30 minutes once a week gets about 26 hours of therapy a year. The same child practicing 10 minutes a day at home adds roughly 60 more hours annually. The math makes the case on its own. Still, an SLP sets the targets, adjusts course when things stall, and notices things a parent watching every day can miss. If your child isn't yet in speech therapy and is past 2.5 to 3 years old with no reliable responses to even basic "what" questions, a formal evaluation is the right next step.
Early intervention services before age 3 are available at no cost in the United States under the Individuals with Disabilities Education Act, Part C. After age 3, Part B of IDEA covers school-based services for eligible children. [10] You don't need a doctor's referral to request an evaluation from your local school district.
For autistic children, autism spectrum speech therapy usually handles question-answering as part of a broader communication plan that includes social pragmatics and AAC when needed. The strategies here line up with what good ABA and speech-language programs already do, but they don't stand in for individualized professional support when a child's needs are complex.
Common mistakes parents make with 'what' questions
A few patterns show up again and again. Asking too many questions in a row is one: three questions back to back overwhelm a child who's still building the skill, so it's better to ask one, wait, respond to whatever happens, and move on. Another is turning questions into a test rather than a teaching moment. "What is that?" asked to check whether your child knows something feels very different to a child than "What is that?" asked inside play and shared attention. Keep it playful; children feel the difference. Harsh or repeated correction is another trap. If a child says "ball" while looking at a balloon, saying "That's a balloon" once is fine. Drilling the correction, showing frustration, or circling back later to quiz them just trains avoidance. Parents also tend to expect generalization to happen on its own, but a child who answers "What is this?" about pictures at the table may not answer the same question about real objects at the grocery store. Generalization has to be taught explicitly, across settings and people, so plan for it. And finally, resist moving up levels too fast: answering 80% of Level 1 object questions doesn't mean your child is ready for "What did you do today?" That question may be months away. Stay at the level where they're succeeding.
Working 'what' questions into the day without formal sessions
Embedding practice in daily routines works better than a separate "therapy time" for most families. At meals, ask "What do you want to eat?" with food visible on the table, or "What is this?" about each item, using forced choices if needed. In the bath, name every object as a question: "What is this?" while holding up the shampoo, answering yourself if your child doesn't. Keep it light. Books are one of the oldest and best-studied language-building routines around (shared book reading, or dialogic reading), so pause on a page and ask "What is that?" about clear, nameable pictures. [11] During pretend or toy play, narrate and question: "What does the cow eat?" "What is in the box?" And outside, point at things on a walk, a truck, a tree, a puddle, and ask what they are. Real places with real objects are an underused language context. Aim for density without pressure: lots of question chances across the day, low stakes on each one, plenty of models and supported answers. That's the setting where the skill grows fastest. If you want structured daily practice alongside your work with an SLP, Little Words (littlewords.ai) offers an AI speech companion that gives parents targeted "what" question activities matched to a child's current level. It won't replace therapy, but it makes consistent home practice easier to stick with.
How do I know if my child is making progress?
This is measurable if you track it. Once a week, run 10 structured "what" question trials at the same difficulty level, and write down how many your child answers correctly without prompting, versus with prompting, versus not at all. A reasonable benchmark: a child at 0 to 30% correct with prompts in week one, moving to 50 to 60% correct unprompted by week four or five of daily practice, is showing real movement. If nothing has shifted after two to three weeks of consistent effort, change the approach rather than pushing harder, and that's a good time to call the SLP. Watch generalization too. Can your child answer the same type of question with a different person, a grandparent or a teacher, in a different place, with different objects? It's normal for generalization to lag acquisition by two to three weeks. But if a child hits 90% accuracy in the kitchen and zero at school, that gap is worth targeting directly.
Common questions about teaching 'what' questions
Most children start answering simple "What is that?" questions between 18 and 24 months, and by 4 to 5 they can handle trickier ones about events, functions, or hypotheticals. If your child is 3 or older and isn't answering basic object-naming questions at all, it's worth getting an evaluation from a speech-language pathologist.
If your child repeats your question back instead of answering it, that's echolalia. It shows up often in autistic children and some late talkers, and it isn't defiance or confusion in the way people usually mean those words. The child might be processing, buying time, or responding the only way they can in that moment. Try fill-in prompts like "This is a..." and model the answer yourself before expecting an independent response. An SLP can help sort out what's driving it.
One technique worth knowing is errorless learning: you set up the question so the child almost can't get it wrong. For "what" questions, that means asking "What is this? It's a..." and supplying the answer yourself if nothing comes within a few seconds. The point is to build a correct retrieval pathway without frustration, then gradually stretch out the wait time so the child has room to answer on their own. Real objects work better than picture cards at first. A snack on the table, a toy in hand, a washcloth at bath time all give richer context than a flashcard, so start there and move to pictures once your child names real things reliably. Skipping straight to pictures can leave a child who nails the flashcard version but goes blank on the same question in real life.
For a nonverbal or minimally verbal child, start with AAC. A choice board with two picture options lets a child answer by pointing or selecting, which builds the habit of responding without requiring speech yet. Treat a pointed or device-based answer exactly as you would a spoken one. The comprehension and response skills transfer even before speech shows up, and an SLP with AAC experience is a real asset here.
"What" questions are the easiest wh-question type, so teach them first. They anchor to things a child can see and touch. "Who" comes next, then "where," then "when," with "why" and "how" last since they require actual reasoning. Don't move on to the next type until the current one is solid, meaning 70 to 80 percent accuracy.
Screens can be part of practice, but only if you're interacting. Passive watching doesn't build vocabulary or question-answering skills on its own. Pausing a short clip and asking "What is that?" about what you just watched together is a fine way to practice, especially with older toddlers and preschoolers.
Timelines vary a lot. A neurotypical child who simply hasn't been prompted much may give reliable answers within two to four weeks of daily practice. A child with autism or a significant language delay might need several months to move through the levels and generalize the skill across people and places. Five or ten minutes of practice every day beats one long session a week.
If your child answers at home but not at school or with unfamiliar people, that's a generalization problem, not backsliding. The skill got learned in one context and hasn't traveled yet. Practice on purpose with other people and in other places: a grandparent, a neighbor, a teacher can all run the same routines. Start with items your child has already mastered in the new setting so they succeed before you raise the difficulty.
Answering "what" questions does show understanding, but it's not the whole picture. A child can understand a word receptively without being able to produce it on demand. Answering taps receptive understanding, expressive retrieval, and the social knowledge that a question calls for a response, so a child can have strong comprehension alongside a real expressive gap.
A few apps include "what" question practice, and quality varies. Look for ones that track your child's level, let you customize vocabulary, and were built with SLP input. Little Words (littlewords.ai) is one built specifically for neurodivergent kids and late talkers, matching prompts to a child's current level. Any app works best with a parent sitting alongside, not handed over for solo use.
As for when to worry: a 2.5-year-old who doesn't respond at all to "What is that?" about something right in front of them is worth a professional look, and so is a 4-year-old who still can't answer "what" questions about events or functions. Neither means something is wrong, but both are past the point where waiting it out makes more sense than asking a professional.
Sources
- ASHA, Typical Speech and Language Development: Developmental milestones for question comprehension and answering in children ages 1 to 5
- ASHA, Late Language Emergence (Practice Portal): ASHA recommends evaluation when caregivers have concerns about language development; late language emergence is a recognized clinical category
- Prizant BM, Duchan JF (1981). The functions of immediate echolalia in autistic children. Journal of Speech and Hearing Disorders, 46(3), 241-249.: Echolalia in autistic children often serves communicative functions rather than indicating absence of comprehension
- Fey ME et al. (2003). Journal of Speech, Language, and Hearing Research: Teaching language skills in a progression from concrete, contextually supported items to abstract produces better outcomes than starting with abstract targets
- Weitzman E, Greenberg J (2002). Learning Language and Loving It. The Hanen Centre.: Self-talk modeling (narrating and answering questions aloud without requiring child response) is a validated naturalistic language facilitation strategy
- Kaiser AP, Roberts MY (2013). Parent-implemented enhanced milieu teaching with preschool children with intellectual disabilities. Journal of Speech, Language, and Hearing Research, 56(1), 295-309.: Real objects and natural contexts significantly improve word retrieval and expressive language outcomes compared to decontextualized practice
- ASHA, Augmentative and Alternative Communication (Practice Portal): AAC-supported communication responses should be reinforced equally to verbal responses; AAC supports rather than replaces speech development
- Lovaas OI (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical Psychology, 55(1), 3-9.: Children with autism may require 20 to 200 massed trials to reach criterion on a discrete skill, depending on complexity and individual factors
- Roberts MY, Kaiser AP (2011). The effectiveness of parent-implemented language interventions: A meta-analysis. American Journal of Speech-Language Pathology, 20(3), 180-199.: Parent-implemented language interventions between therapy sessions significantly improve outcomes compared to therapy alone
- U.S. Department of Education, Individuals with Disabilities Education Act (IDEA): Part C of IDEA provides free early intervention for children under age 3; Part B covers school-based services for eligible children ages 3 and older
- Whitehurst GJ et al. (1988). Accelerating language development through picture book reading. Developmental Psychology, 24(4), 552-559.: Shared book reading (dialogic reading) with question prompts about pictures is one of the best-studied naturalistic language-building routines
- American Academy of Pediatrics, Developmental Surveillance and Screening: AAP developmental surveillance guidelines include language milestones and recommend referral when milestones are not met