The mental tax of writing English as a second language
You speak French. Or Spanish, German, Japanese, Mandarin. Whatever your native tongue is, English is the language you have to write in for work. Emails to clients. Slack messages to coworkers. LinkedIn posts. Documentation. README files.
And every single time you sit down to write one of those messages, the same thing happens: you think the sentence in your native language, you translate it in your head, then you type it in English. Three steps for one sentence. By the end of a workday, you are exhausted. Not from the work. From the translation.
This is the silent tax non-native English speakers pay every day. It is invisible to colleagues. It does not show up in any productivity report. But it is real, and it shows up as mental fatigue, slower output, and the constant low-grade anxiety of sounding non-native.
The good news: voice dictation, done right, can fix most of this. Not all of it. But enough that you stop noticing the tax.
Why typical dictation tools do not solve this
Most dictation apps for Mac transcribe what you say in the language you speak. Useful for native speakers. Useless for the non-native English writer who needs the output to be in English, not in their native tongue.
Some tools, like SuperWhisper, offer a translation mode. But it is buried in a Pro tier, configured through custom modes, and only translates to English. You cannot quickly switch the output language. You configure it once and it stays.
Others, like Wispr Flow, do excellent multilingual transcription with auto-detect and code-switching. They handle the case where you mix two languages mid-sentence. But they do not translate. If you speak French, you get French text. End of story.
Apple's built-in dictation does not translate either. Google Translate has voice input but cannot type the result into the app you are working in. You end up with a clipboard dance: speak into one app, copy, paste into another. Two minutes of friction every time.
What actually works: speak in your language, write in any other
The workflow that actually fixes the mental translation tax looks like this:
- You think in your native language. No translation step.
- You speak in your native language. Naturally, fluently, without searching for English words.
- You hit a shortcut to switch the output language to English. One keystroke.
- Clean English text appears wherever your cursor is. Gmail, Slack, Notion, VS Code, anywhere you can type.
That is what Scrybapp does. You dictate in French, you get English. You dictate in Japanese, you get Spanish. The output language is a setting you change with a shortcut, not a mode you configure once and forget.
The bidirectional part matters. Most translation tools assume one direction (your language to English). But sometimes you need the reverse: writing back to a French client in French while you naturally think in English now after years of expat life. Or replying to a German customer in German when your German is rusty. The shortcut handles both.
How to set this up on your Mac
The setup takes three minutes:
- Download Scrybapp and install it (one-time payment, no subscription).
- Set your input language to your native tongue. This is the language you speak when you dictate.
- Set your output language to English. This is the language that ends up typed.
- Bind a shortcut to "switch output language" so you can change it on the fly when you need to write in a third language.
- Press your dictation shortcut, speak naturally, release. Clean English text appears in whatever app you are using.
That is the entire workflow. No cloud, no internet required, no subscription. Translation happens locally on your Mac using OpenAI's Whisper model running on Apple Silicon.
Why local processing matters when you are translating
Cloud translation services log your audio. Some of them keep it for years. If you are dictating client emails, business plans, internal feedback, or anything sensitive, that is a real privacy issue most people do not think about until something leaks.
Scrybapp processes everything on-device. Your voice never leaves your Mac. Your translated text never touches a server. This matters more when translating than when transcribing, because the things you say in your native language are often the most personal: complaints, drafts of difficult emails, conversations with yourself about a tricky decision.
For a deeper look at why local processing is becoming the standard, read our local vs cloud speech-to-text comparison.
Use cases this unlocks for non-native English speakers
- Client emails in English. You think in your native tongue, the email comes out in clean English. No more rewriting awkward phrases.
- Slack messages. Quick replies stop being a 30-second deliberation. You speak the answer, you send it.
- LinkedIn posts. You think out loud in French, you publish in English without losing your authentic voice. The translation keeps your tone, not just your words.
- Documentation and README files. You explain a feature in your most fluent language, the docs land in English for the international team.
- Translating drafts you already wrote. Speak a draft in your native language out loud, switch the shortcut, paste the English version. No round trip to DeepL or Google Translate.
- Cold outreach in a third language. Reaching out to a Spanish prospect when you only speak French and English? Speak in French, set output to Spanish, send.
The accuracy question (and an honest answer)
Machine translation is imperfect. Idioms, slang, technical jargon, cultural nuance, all of these still trip up local AI translation. You will get the gist right around 90 to 95% of the time. The remaining 5 to 10% needs editing.
Honest take: this is not a tool that replaces a fluent English writer. It is a tool that gets you 90% of the way there in 10% of the time. You then spend a minute polishing the result instead of an hour drafting it from scratch.
For most professional use cases, that trade is overwhelmingly worth it. The mental energy you save by not translating in your head is energy you can spend on the actual work.
Where it works less well: highly idiomatic content (humor, marketing taglines, song lyrics), legal language with strict terminology, and very technical fields where the right English term is not a direct translation. For those, treat the output as a first draft and polish.
What about your accent
Whisper, the model Scrybapp uses, was trained on speakers from dozens of countries with all kinds of accents. We have seen it work well on French, Spanish, German, Italian, Portuguese, Japanese, Mandarin, Korean, Arabic, Hindi, Vietnamese, Polish, Dutch, and many more. Heavier or rarer accents may need slightly slower speech to keep accuracy high.
Bottom line: your accent stops at your keyboard. The model does not care how you sound, it cares what you say.
If you have ever felt self-conscious about dictating because of your accent, this is the tool that removes that friction. Nobody hears your voice. Only you and your Mac.
How this fits with the existing tools you already use
Scrybapp does not replace your text editor, your email client, or your messaging app. It works inside them. The output language switch is a single shortcut you press before dictating, and the translated text gets typed directly where your cursor is.
So your workflow stays the same. You are still in Gmail, Slack, VS Code, Notion. You just type with your voice instead of your keyboard, and the output language is whatever you set it to.
If you currently use DeepL, Google Translate, or ChatGPT to translate drafts back and forth, this collapses that workflow into a single step. No more copy-paste, no more switching tabs.
Stop paying the translation tax
If you write in English for work and your native language is something else, the workflow above gives you back the hours you spend translating in your head every week. You speak in your tongue, you write in any other.
Get Scrybapp, set up the output language shortcut, and notice how much less tired you feel at the end of a workday. The translation tax is invisible until you stop paying it.
Related reading: Multilingual voice typing on Mac, Speak Spanish, write English, How to dictate in multiple languages on Mac.