Hapaxes - words that occur very rarely or only once in the text - offer an interesting window on the problem (see also previous posts: ضريع, قسورة). Apart from the name Yūnus (Jonah) itself, four words are attested in the Qur'ān only within accounts of Jonah. The oldest attested form of his name is Yônāh, which in Greek yields Iônas (ιωνας) in the nominative (the -s is a widespread Indo-European nominative singular suffix); the final s in Yūnus thus suggests that the audience's knowledge of Jonah came in part via Greek intermediaries at some remove. "Fish" is normally ḥūt حوت in the Qur'ān, including in accounts of Jonah (Standard Arabic samak سمك is unattested in the text), but in 21:87 Jonah is alluded to as ḏā n-nūn ذا النون "he of the fish", the only occurrence in the Qur'ān of the Aramaic loanword nūn. The fish swallowed (iltaqamat التقمت) Jonah in 37:142; the only other mention of swallowing in the Qur'ān uses a word much better attested in modern Arabic dialects, balaʕa بلع (11:44:3). After praying to God for release, he is then cast out onto the shore, for which both 37:145 and 68:49 use the more specific term ʕarā' عراء, ie barren land. Eventually God causes a gourd - yaqṭīn يقطين - to grow over his head; this is the only Qur'ānic mention of the plant in question.
Compare the relevant terms in various early Semitic versions of the Book of Jonah:
|Jonah||fish||swallow||land sp.||plant sp.|
|Arabic||Yūnus يونس||ḥūt حوت / nūn نون||iltaqama التقم||ʕarā' عراء (barren land)||yaqṭīn يقطين (gourd)|
|Hebrew||Yônāh יוֹנָה||dāḡ דָּג||bālaʕ בָּלַע||yabbāšāh יַבָּשָׁה (dry land)||qîqāyôn קִיקָיוֹן|
|Babylonian Jewish Aramaic||Yônāh יוֹנָה||nūnā נוּנָא||blaʕ בְּלַע||yabbeštā יַבֶּשׁתָּא (dry land)||qîqāyôn קִיקָיוֹן|
|Syriac||Yawnān ܝܘܢܢ||nūnā ܢܘܢܐ||blaʕ||yaḇšā ܝܒܫܐ (dry land)||qar'ā ܩܪܐܐ (gourd)|
|Geez||Yonas ዮናስ||ʕanbari ዐንበሬ (whale)||wəxṭä ውኅጠ||mədər ምድር (land)||ḥamḥam ሐምሐም (gourd)|
One immediately notices that none of them match the Qur'ān as a whole at all well. For "Jonah", only Geez (Ethiopic) offers a similar Greek-influenced term, contrasting with the obvious Aramaic source of nūn for "fish". For "swallow", the Hebrew and Aramaic/Syriac versions all use a word whose direct cognate - balaʕa - is attested elsewhere in the Qur'ān, and is very familiar in Arabic; why then does the more vivid term iltaqama (something like "take in as a morsel") appear? For the land onto which Jonah is cast, the Qur'ān twice uses a specific term incorporating a detail absent from any of these versions of the Book of Jonah, all of which use a generic term for "dry land" or even just "land"; why is this used rather than 'arḍ or even the cognate yābisah?
The conclusion seems obvious: none of these translations were at all prominent for the Arab audience to whom the Qur'ān was first addressed. Whatever its distant roots may have been, the account of Jonah they knew best was something orally transmitted in Arabic, and not directly based on any one of these.