A show's language field is useful metadata, but it cannot explain every episode. A podcast may mix languages, publish occasional interviews in another language, or use an incomplete language tag.
If your app offers a language filter, describe what it filters. “Shows marked as English” is a more precise promise than “Every episode is in English.” Keep the filter optional when possible.
Separate interface language from content language
A Turkish interface does not mean the listener wants only Turkish audio. Store interface preferences and podcast discovery preferences separately. Let a listener search an original title without translating the text into the app's current language.
Use documented language values from the provider. Do not assume a country code identifies a language, or that one language belongs to only one country. If you simplify choices into readable labels, keep the original data in your model.
Missing language should remain a recognizable case. You can include such records in an unrestricted search without pretending they match a narrow filter. Explain that removing a language filter may uncover additional shows.
Test a multilingual person, not only a locale
Try a listener who uses a German interface, searches an English title, and follows a Japanese show. Their saved choices should survive navigation and a later interface-language change.
Also test accented characters and a query entered through a composing keyboard. Search behavior should not quietly strip meaningful text to fit an English-only assumption.
Language metadata works best as one part of discovery, alongside descriptions and previews. Giving listeners control is more useful than treating a single catalog field as a complete account of what they will hear.