MEWSIGNALCat communication board
Stored on this device
✦ A cat-centered communication experiment

Make room for their signal.

A paw-sized soundboard, a gentle learning guide, and an observation journal. MewSignal helps you notice patterns without pretending a tap is a translation.

Paw Pad · cat-facing view

Miso's board

Voice on
Ready. A touch is recorded as an event; review its context before drawing conclusions.

One broad target · calm colors · no flashing. Touch checks confirm a screen event only; they do not verify a cat's intent. Screen response varies by device and paw.

Learn · voluntary practice

Teach one cue through a routine you already share.

No pressure
01
Choose one useful, low-risk signal.

Pick something your cat already asks for in a clear routine, such as play. Keep essential needs available without a press.

02
Use one label and one familiar voice.

Say the same short word in the same context. A caregiver recording is more recognizable than synthetic speech; keep volume comfortable.

03
Model the button, then follow through.

Use “Model this cue” in the caregiver workspace, say the word, then do the safe matching activity. This plays the cue without logging a cat press. Let the cat observe; never move its paw for it.

04
Make the cat's choice easy to read.

Place the tablet on a stable surface. Check touch detection with one large card. Allow approach and exit; stop if your cat disengages or seems unsettled.

05
Add another signal only after the first is steady.

Look across different days and situations. Repeated presses alone do not prove understanding; compare context, response, and whether your cat chooses to return.

Board setup

Choose what your cat sees and hears.

Voice starts off. Preview it yourself at a low level before using it near your cat.
Recorded voice and browser speech; device loudness varies.
Preview away from your cat first. Individual sound tolerance varies; stop if your cat avoids or seems uncomfortable.
Short, abstract cues only. The device’s own volume remains important.
Journal · caregiver observations

See the moments around a press.

0Observed taps
0Need context
0Session records
Interpretation stays with the human. Mark what you observed, including uncertainty. MewSignal does not infer intent, translate thoughts, diagnose health, or use these counts to grade your cat.
Button setup and saved voice clips are not removed.
Field Guide · evidence and limits

A useful aid can be honest about what it does not know.

Research-informed · not validated as translation
What this prototype is: a customizable soundboard, a voluntary practice aid, and a local record of what happened around a button press.
What it is not: a cat-language translator, a diagnostic device, a validated feline AAC protocol, or proof that a press carries a specific meaning.
Feline learning · 2017

Clicker training for shelter cats

A 100-cat study found improved performance on four cued behaviors after 15 brief sessions. Food motivation and temperament related to some outcomes. It supports trainability and individual variation—not button-word comprehension.

Read the peer-reviewed study ↗
Word association · 2024

Rapid picture–word association

Cats detected changed picture–speech pairings, but the same paper found no significant switch effect with electronic sounds alone. This short-term task supports possible association, not stable button meanings or long-term word learning.

Read the peer-reviewed study ↗
Animal–computer interaction · 2025

Tablet play in a cat café

A five-month study with 16 cats compared a video-based tablet experience with toys. It highlights the need to attend to feline behavior and autonomy; it does not validate touchscreen soundboards.

Read the ACI conference paper ↗
Active research

Soundboard questions remain open

UC San Diego's Comparative Cognition Lab describes ongoing work on soundboards with dogs, cats and other companion animals, including owner reports and controlled studies. The live research is still investigating what animals learn and communicate.

Explore the research program ↗
Feline welfare · AAFP/ISFM

Choice, predictability and control

Environmental-needs guidance emphasizes positive, predictable contact that lets cats initiate, shape and end interaction. MewSignal treats walking away as a valid end state, not a failed session.

Read the feline environmental-needs guidelines ↗
Human–cat cues · 2023 pilot

Visual and vocal cues together

A small café study coded 12 cats and found approach latency differed across visual, vocal and combined human cues. It cannot establish home preferences or soundboard meaning; it is a reason to keep the visual board useful even when voice is off.

Read the peer-reviewed pilot study ↗
Interface principle

Color is never the only cue

Cat color perception differs from ours. This board also uses large separated targets, consistent words and distinct icons. Check visibility and paw response on the actual device; do not assume every tablet senses every cat's paw.

Read a feline vision study ↗
Product boundary

Observation before automation

Logs are kept on this device. No microphone runs in the background, no behavior is auto-labeled, and no press is turned into an AI-generated meaning. A caregiver can mark a moment supported, uncertain or accidental.

Open the local journal →
First household trial · keep it descriptive
  1. Begin with one large, silent cue for a low-risk routine such as play; secure the device and use the touch check if needed.
  2. Let your cat approach, touch, pause or leave without prompting or physically guiding a paw. Keep food, water, litter, medication, shelter and care available as usual.
  3. Note what happened just before and after a touch. Repeat on different days before considering a pattern; one household trial cannot establish what a button means.
Start safely: use a stable device with one large target; let the cat approach and leave; pair a consistent word with an ordinary routine; respond promptly when safe; keep all essentials available; stop if the activity creates frustration. If a behavior changes suddenly, consult a veterinarian rather than relying on a button log.
Back to DV Labs demos