A visionOS app where you write Chinese in the air and feel a haptic response is a genuinely appealing demo. Before treating it as a way to build retention, two honest distinctions matter. A mid-air buzz is not the friction of pen on paper, and air-writing is a different motor skill from handwriting. Underneath both is the deeper point the demo glosses over: feeling and recognizing a character is not the same as recalling it.
Haptics simulate a tap, not a stroke
Real handwriting feedback is continuous: the resistance of the tip, the friction of the surface, the subtle drag that tells your hand how a stroke is going. Current mid-air haptics deliver discrete cues, a buzz, a tap, which approximate touch coarsely but do not reproduce that continuous physical signal. So a haptic API can confirm you crossed a point in space; it cannot give your hand the felt resistance of writing. That gap is why the simulated feedback supports immersion more than it builds the motor pattern, the same limit seen when comparing stylus hovering to VR finger-tracing.
Air-writing is the wrong scale of movement
Even setting haptics aside, writing a character in the air is gross motor: large arm and shoulder movements through a big volume. Handwriting is fine motor: small, precise finger movements at the size of a fingernail. Motor learning is specific, so practicing the big-volume version does not fully transfer to the small one you use with a pen. That is the recurring caveat behind writing massive virtual characters and headset hand-tracking demos: impressive movement, mismatched scale.
Recognition is not recall
Here is the deeper issue the immersion obscures. Feeling a character form in front of you, with a guide and a buzz, is a cued, recognition-like experience. Writing it later on paper is uncued recall, producing it from nothing. They are different, and only the second is what you actually need. The testing effect shows retrieval builds memory far more than re-exposure, classic work found physically handwriting shapes improved later recognition, and for Chinese, handwriting beats typing for learning through the motor and language networks it recruits. A guided, haptic air-trace engages little of that; from-memory production engages all of it.
Simulated feedback versus real production
| visionOS haptic air-writing | Real-surface from-memory practice |
|---|---|
| Discrete buzz, not friction | Continuous real feedback |
| Gross motor, big volume | Fine motor, true size |
| Guided, recognition-like | Uncued recall |
| Supplement for immersion | The retention mechanism |
The honest read: spatial haptics are a fun supplement, not the engine of memory.
A plan if you have a headset
- Use haptic air-writing for novelty and first exposure only.
- Do the real learning on a surface, at true character size.
- Produce each character from memory, not by guided air-tracing.
- Check stroke order and structure on the real attempt.
- Space the repeats so production, not recognition, is trained.
How Hanzi Write Practice fits
Hanzi Write Practice is candid about all of this: it is not a visionOS haptic app, and it does not promise that a mid-air buzz builds memory, because the mechanism is from-memory production on a surface that matches real writing. It hides the character, you produce it at true size, and it checks stroke order and structure with spaced repetition. If spatial computing later earns a role as an on-ramp, the retention still comes from the fine-motor, recall-based core, not the haptic novelty. The app is in early access.
Bottom line
A visionOS haptic air-writing app adds a buzz, not the friction of pen on paper, trains gross rather than fine motor, and is recognition-like rather than recall, so it is a supplement, not the mechanism. Retention comes from producing characters from memory on a real surface. Hanzi Write Practice trains exactly that, and it is in early access, so join the list.
Frequently asked questions
Does haptic air-writing on visionOS help you remember characters?
It can aid engagement, but it is a supplement, not the mechanism. Haptic feedback in mid-air does not reproduce the friction of pen on a surface, and air-writing uses gross motor movement rather than the fine motor of handwriting. Retention comes from producing characters from memory, which a real-surface, from-memory tool like Hanzi Write Practice trains directly.
Is recognizing a character the same as being able to write it?
No, and conflating them is the core mistake. Recognition is cued: the character is in front of you and you confirm it. Recall is uncued production from memory, which is what writing requires. You can feel and recognize a character in an immersive demo and still be unable to produce it on paper.
Can simulated haptics replace the feel of real handwriting?
Not really. Real handwriting feedback comes from continuous friction, resistance, and contact between a tip and a surface, which current mid-air haptics approximate only coarsely. The simulated cue can support a sense of touch, but it is not the same physical signal, so it does not build the same motor pattern.
What actually builds Hanzi retention?
Producing characters from memory, with stroke-order feedback, spaced over time, ideally on a real surface so the motor pattern matches writing. That is the evidenced path to retention. Hanzi Write Practice is built around from-memory production on a device you actually write on, rather than mid-air gestures.
Tempted by the headset demo? Join early access and train the production that actually sticks.