Dyslexia does not switch languages the way your keyboard does

Problem

Most dyslexia-friendly features in software were designed around English, so dyslexic users who live in two languages get half a solution. Dyslexia shows up differently in each writing system. In English it looks like spelling and decoding errors. In Farsi, everyday text drops vowel marks, so reading means sounding out consonants. In Mandarin there is no alphabet: typing means picking the right character from a list of look-alikes. A bilingual person with dyslexia switches between these systems dozens of times a day, and every switch resets the keyboard, the autocorrect and the predictions. Nobody owns that moment.

Question

When bilingual adults with dyslexia switch languages on their phone or computer, where do things break, and what workarounds have they built? It names one observable moment, and asking about workarounds turns every answer into a design opportunity. The scope held to English paired with Farsi and with Mandarin, chosen because the three scripts fail in different ways; to adults who use both languages on a device most days, with self-identified dyslexia counting; and to messaging, search, forms and reading. The focus is the switch, not dyslexia in general.

Method

This phase is secondary research built on peer-reviewed studies only. I had no access to participants, so I worked from published evidence and was explicit about where it stops. I reviewed the literature on dyslexia in Persian and Chinese scripts, including brain-imaging work showing the two use different reading pathways, and studies of adults with dyslexia reading, spelling and typing in a second language. I mapped what each study covers and what it does not: almost all of it is lab reading and spelling, mostly with children, and none covers bilingual adults switching languages on a device. From that I built three provisional personas, each a pattern rather than a person, with every trait tagged sourced or hypothesis so nothing invented is presented as data. Then I extracted findings and recommendations, with confidence rated low until interviews confirm them.

Personas

Three patterns, each breaking at a different step. Names are placeholders; the full trait tables, with every line tagged sourced or hypothesis, are on the personas and sources page.

Findings

Four findings, all low confidence until interviews confirm them, and that is stated rather than hidden. Brain-imaging research shows Chinese and alphabetic reading impairments involve different regions, and early bilingual reading changes how dyslexia shows up in adults, yet products still treat a dyslexic user as one account-level setting. The last finding is a gap rather than a result: adults with dyslexia are known to type more slowly and less accurately, but no peer-reviewed study covers bilingual adults on their devices. That gap is the project's main contribution.

Recommendations

The first two are cheap and help every bilingual user. The script-specific ones need validation with real users first, which is why the last recommendation is research rather than a feature.

Pieces

Findings are easier to believe once you have felt them, so the research also ships as three pages of its own: two to try, and one with the full evidence.

• Feel the switchThree two-minute trials that put you inside the problem: read without vowels, pick a character, fight autocorrect.start here if you have 2 minutes
• Friction mapOne message from first tap to send. Nine stops, each tagged by persona, evidence and fix to test.start here if you want the analysis
• Personas and sourcesThree provisional personas with every trait tagged sourced or hypothesis, plus the nine peer-reviewed sources.the evidence base

Next

A six-to-eight person interview study, two or three per language pair, using the hypotheses as the interview guide. It would produce validated personas, a friction map of the switch moment checked against real phones, and prioritised recommendations. My background is in design, support and teaching, and all three taught me the same thing: people rarely tell you where they are stuck, but they always show you through the workarounds they invent. If you are a Farsi–English or Mandarin–English speaker with dyslexia, or know someone who is, I would like to talk.