Should Arabic content be translated, adapted, rewritten, restructured, edited or written from scratch?

The six treatments in the Arabisation Framework are options, not levels of quality. Translation is not the basic choice and rewriting is not automatically the better one. A rewritten page can lose accuracy that a disciplined translation would have kept, and an expensive rebuild can solve a problem that careful editing would have solved faster.

The useful principle is simple: the right treatment is the least intervention that makes the communication work properly in Arabic. Anything less leaves the failure in place. Anything more spends effort and risk on a problem that was not there.

The framework asks nine questions across three levels, and the treatment falls out of the answers rather than being decided in advance.

• Communication: Purpose, Audience, Context. What should the content achieve, who receives it, and where does it appear?
• Arabic: Meaning, Expression, Voice. Is the substance right, does it read as Arabic, and does it sound like the right person or organisation?
• Project: Structure, Terminology, Coherence. Is it organised for Arabic readers, are the language decisions deliberate, and does the whole body of content hold together?

One rule saves most of the argument later: locate the failure first, then pick the treatment that reaches it. If you cannot name which of the nine is failing, you are not yet ready to choose.

Each treatment answers a different kind of failure. Read these as rules of thumb for a single piece of content, or for a single section within a larger project.

• If purpose, audience, context, structure and voice all survive the move into Arabic, then translate. Signal: the English works and nothing essential about it is tied to an English-speaking reader.
• If the central message survives but a reference, example, tone or expression does not carry, then adapt. Signal: readers would understand the words but the effect would be weaker or slightly off.
• If the meaning should stay but the Arabic has to be rebuilt sentence by sentence, then rewrite. Signal: every fix you attempt pulls the next sentence with it.
• If the order, hierarchy, headings or format are part of the problem, then restructure. Signal: no sentence is clearly wrong yet the reader has to work to find the point.
• If usable Arabic already exists and needs correction of expression, voice, terminology or consistency, then edit. Signal: the draft is close and the changes are local rather than structural.
• If starting from an English draft would constrain the argument or the voice, then write in Arabic from scratch. Signal: the strongest Arabic version would not resemble the English one.

Where two rules seem to apply, choose the lighter one and test it on a representative section. If the lighter treatment leaves the failure visible, escalate deliberately rather than by default.

These are generic cases, not client work, and they show how the same nine questions produce different answers.

A corporate web page describing a service. Meaning is stable and the purpose is the same for both audiences, but headings and calls to action often read as imported. The usual treatment is translation for the body copy with adaptation of headings, buttons and metadata, and restructuring only where the page order assumes an English reader's journey.

An executive op-ed. The purpose is to establish a point of view, so voice dominates the judgement. If the English piece already carries a real argument, rewriting in Arabic is often right. If the argument itself is thin or was assembled for a different market, original Arabic writing from a briefing with the executive is the stronger route.

A campaign headline. The failure is rarely meaning alone. It is usually effect: rhythm, wordplay, cultural fit and the promise the line makes. Adaptation is the default, and original Arabic writing is appropriate when the concept depends on an English construction that has no Arabic equivalent.

A press release built from a brief rather than an English draft. Purpose, audience and context are clear, and there is no source text to preserve, so writing directly in Arabic is usually more efficient and often produces a stronger result. Attempting an English draft first tends to import a structure that Arabic newsrooms may not expect.

An AI-generated Arabic page. Expression is often fluent while meaning, voice and terminology drift. Verify substance first. If the facts and argument hold, editing may be enough. If the page sounds like nobody in particular or invents specifics, rewriting is usually the more honest treatment, and light editing is often the expensive mistake.

A multi-page website with terminology inconsistency. Individual pages may each be acceptable, so this is a project-level failure. The treatment is editing plus terminology decisions applied across the whole site in one pass, with restructuring reserved for the pages where the organisation, not the wording, is the problem.

Most real projects use more than one treatment at once. A website can be translated in its body copy, adapted in its calls to action, restructured on two pages and edited across the whole for terminology. A campaign can carry a translated fact sheet, an adapted headline and an original Arabic launch statement.

Deciding treatment section by section is not indecision. It is what keeps the work proportionate: the parts that travel well are left alone, and effort goes where the Arabic actually fails.

Before committing to a treatment, ask three questions in order. Which of the nine points is failing? What is the lightest treatment that reaches that point? What would break if the treatment were one step lighter?

If the third question has no clear answer, the plan is probably heavier than the content needs. If it has an obvious answer, the choice is defensible, and that is the standard worth holding: not the most thorough treatment available, but the one the communication can be shown to require.