Letters that won't join, text reading the wrong direction, or a PDF where everything looks scrambled — this is a known, common software problem, not something wrong with your Arabic text itself.
Diagnosed and explained by Ata Rafi, Founder and Chief Formatter at Format Fix, from direct experience fixing this exact problem.
Request A Free QuoteThis almost always comes down to one of two things: the font itself doesn't include the shaping data Arabic needs, or the software displaying the text doesn't know how to apply it — even if the font supports it.
Arabic reads right-to-left, but a page can mix Arabic with English words, numbers, or punctuation — and each of those pieces has its own direction. Software has to apply a set of rules (the Unicode Bidirectional Algorithm) to work out the correct reading order for that mix. When it's applied incorrectly or not at all, whole phrases can appear reversed, or Arabic and English can swap places on the line.
This is the same underlying issue behind Arabic-English tables or captions that look scrambled — the layout tool is treating the line as one direction when it actually contains two.
These are the specific, recurring situations where this happens in real book projects — not abstract technical edge cases.
Most mainstream design and publishing software was built with Latin-script languages as the default. Arabic's letter-joining and right-to-left requirements are a separate layer of support that not every tool includes.
A standard, Latin-script-first edition of a layout program — Arabic characters may display, but shaping and right-to-left handling are often incomplete or missing entirely.
A dedicated edition or Arabic-aware tool built with full shaping and right-to-left support — the same underlying document behaves correctly.
This is exactly why the same Arabic text can look correct in one program and broken in another, and why it can look fine while typing but break the moment it's placed, exported, or opened elsewhere. Format Fix's typesetting work is done entirely in Arabic-aware tools built for this — see Arabic Editorial Formatting for the structural process this sits underneath.
In the vast majority of cases, the Arabic characters themselves are correct and intact — the display or output step is what's failing to render them properly. This is different from a rarer, separate problem: text that shows up as boxes, question marks, or unreadable symbols, which usually means the file's character encoding doesn't match what the software expects, rather than a shaping or direction failure.
In either case, the manuscript itself doesn't need to be rewritten — it needs to be handled by a tool and workflow built for Arabic from the start. See 200+ Arabic, Urdu, and bilingual books formatted correctly for what that actually looks like in practice.
This is standard practice on every Format Fix project, not an extra step — correct shaping and direction are checked at layout, at proofing, and again at final export.
Writers drafting in Word or a general word processor who see the text look fine while typing, then break once moved into layout.
Files arriving from different authors, translators, or typists, each produced in different software with inconsistent Arabic support.
OCR'd text from a scanned Arabic book can lose its correct letter-joining behavior in the conversion, needing the same shaping and direction fixes as any other broken file. See Arabic Typing Services for manuscript digitization instead of unreliable OCR.
The font or software isn't applying Arabic's contextual letter shaping, so every letter falls back to its standalone form instead of joining into a flowing word. If this is happening in a file headed to KDP, see Common Arabic eBook Formatting Mistakes That Get Rejected by KDP for how this specific issue causes rejections.
The software isn't correctly applying the bidirectional reading-order rules Arabic needs, especially where Arabic mixes with English words or numbers on the same line.
Copying between programs can drop the shaping and direction data during the paste, even when both programs individually support Arabic. This is a well-documented, recurring problem, not a one-off mistake. See Fixing Arabic Text Direction in InDesign for what InDesign's RTL settings actually involve, or Arabic Typesetting Software Explained for how InDesign ME and the World-Ready Composer differ.
That's a different, rarer problem — a character-encoding mismatch, where the software is reading the file with the wrong text encoding rather than failing to shape or reorder correctly-encoded Arabic. If it's an actual question mark placed or shaped incorrectly rather than a garbled character, see Arabic Punctuation Placement Problems for how the Arabic question mark is supposed to behave.
Almost never. In the vast majority of cases, the underlying Arabic text is intact — the display or export step is what's failing, not the text itself.
The export step didn't carry the right-to-left and shaping settings through, so the file looks correct in the design program but breaks once opened as an EPUB.
It can almost always be fixed without retyping — the underlying text is usually intact, and the fix is applying correct RTL-aware typesetting rather than re-entering the manuscript. See Request A Free Quote to have a broken file diagnosed.
