What is code-switching? It is the habit of alternating between two or more languages within one conversation or sentence, as bilingual speakers often do. For speech recognition it is a hard case, because a system tuned to one language may misspell the words from the other.
Transcribe a recording
Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.
Open @potok_now_bot in Telegram
Files up to 2 GB · export to TXT, MD, DOCX, PDF, SRT and VTT · speaker labels and timestamps · about 30 languages
Key facts
- Switches can happen between sentences or in the middle of one, and the second case is the harder one to transcribe.
- A recogniser usually picks one language for a stretch of audio, so borrowed words may come out as similar-sounding nonsense.
- Names, product terms and quoted phrases in another language are the most common triggers in business meetings.
- Splitting a long mixed recording at the points where the language changes can help, if you know where they are.
What it looks like
[00:00:05] Speaker 1: Vamos a revisar the budget antes del almuerzo.
[00:00:09] Speaker 2: Claro, and then we send the report to the client.
// one sentence, two languages:
// Spanish frame + English noun phrase -> intra-sentence switch
More
See the full list of languages, the tools and the glossary.
Frequently asked questions
Is code-switching the same as an accent?
No. An accent is how one language is pronounced. Code-switching means actually using words or phrases from a second language while speaking the first.
Why do transcripts go wrong when people switch languages?
The system expects one language at a time. When a second one appears, it forces the sounds into words of the first, and the result is garbled.
Can I get both languages written correctly?
Often only partly. Check the switched passages by ear, and fix them in an editable export. Short, well-known foreign words are usually recognised better than long phrases.
Does the Potok bot turn the second language into the first?
No. The text stays in the languages as spoken and is not converted. Mixed passages need a manual check.
How should I record a mixed-language conversation?
Speak clearly, avoid overlap, and keep each speaker near the microphone. Clean sound helps most when the system has to guess which language is in use.
Sources
Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.
- Code-switching (Wikipedia), Wikipedia · checked
- Language identification (Wikipedia), Wikipedia · checked
Try it on your own recording
Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.
Open @potok_now_bot in Telegram