I know about Persian as an example.

For example, we have "خانه" (transliterated as Khaneh, which means home) for the written form. But we say "خونه" (Khooneh) for the spoken form in conversations.

Or in another example, we have "هنوز" (Hanooz, which means still/yet) for the written form; while its spoken form is "هنو" (Hanoo).

My question is which language has the most distinct written and spoken forms?

  • 2
    In English we have 'Colonel' -> [kəːnl̩] – Mellifluous Apr 19 at 18:51
  • Are you looking only for living languages, or also dead ones? If the latter I nominate the Hittite and Pahlavi scripts, with their abundance of heterograms. – Draconis Apr 19 at 18:54
  • ‘Distinct’ in what way? Some languages have very regular writing systems, but the written form uses different grammar and vocabulary than the spoken form (e.g. Welsh). Other languages are the same when written or spoken, but it is difficult to predict the spoken sounds from the written letters or vice versa (e.g. English). Which case are you asking about? – bradrn Apr 20 at 0:22
  • I would say that Japanese has the most distinct written form. It uses several varieties of Chinese characters, two syllabaries, the Latin alphabet, and Arabic numerals, all used for different purposes, simultaneously. No other writing system comes close in complexity, and the fact that Japan has such a high literacy rate is a phenomenal achievement which has echoes throughout the whole culture. – jlawler Apr 20 at 15:49

The winners are probably writing systems with heterograms: words written in one language but spoken in another. We actually have a few of these in English, such as the abbreviation "e.g." which stands for Latin exemplī grātiā but is usually read aloud as the English equivalent "for example".


  • In Hittite cuneiform, the written form a-na DIŊIR-lim represents Akkadian ana ilim "to the god"; the actual pronunciation was siuni
  • In Pahlavi, the written form LḤMʔ represents Aramaic laḥmā "bread"; the actual pronunciation was nān

Heterograms that actually spell out pronunciations like this are fairly rare; more common are logograms borrowed from one language to another, like Sumerian signs in Akkadian, or Chinese characters (kanji) in Japanese. These are also sometimes called "heterograms", but they don't fit your question in the same way: the connection between 山 and Japanese yama isn't any weirder or less intuitive than the connection between 山 and Mandarin shān, for example.

  • The first Hittite I ever studied was actually mostly heterograms (you could tell because the transcripts used CAPITALS for Sumerian or Akkadian glyphs), but it sure was a common formula. GAL, LU-GAL, LU-GAL GAL LU-GAL, LU-GAL KUR-URU ha-a-ti. The last one being actual Hittite, the rest being "King, Great King, King of Kings, Great King of Ha:ti". Those were the days. – jlawler Apr 20 at 15:57

Tibetan is probably the language written in an alphabet, where the gap between actual spoken phonetics and what's written is just enormous. Many letters are not pronounced and those which are pronounced are often different from what one would expect.

  • This is true of Lhasa/Ü-Tsang Tibetan but not of Amdo Tibetan, at least according to my Tibetan friend from Qinghai who was educated in both. Even then the consensus among experts is that it's still not as bad as English. – hippietrail Apr 23 at 3:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.