A UK hand-hygiene poster, a Hausa audience, and a "verbatim" grant clause
An international NGO is launching a household hand-hygiene campaign with a single English master creative developed in the UK: a poster showing a mother and child at a white ceramic basin with a chrome tap, running water and liquid soap; the slogan "Clean Hands, Safe Lives", chosen because it alliterates; a caption explaining that hands must be washed "with soap and running water for 20 seconds"; and a line thanking the donor. The campaign runs in northern Nigeria in Hausa, with a second variant for urban Kano in Nigerian English. Field staff report that many target households have no piped water and use a kettle or a tippy-tap; that the poster's bathroom looks nothing like local homes; and that "20 seconds" means nothing without a reference point. The donor's brand team requires a back-translation of all copy into English, and the grant agreement states that "key messages must be reproduced verbatim in all languages". The launch is in five weeks and there is no budget for a second photo shoot.
1. Sort the master creative into what can be transferred, adapted within translation, and transcreated - and justify each decision.
Read the worked answer
Transferable. The behavioural core: wash hands with soap, with water that flows over the hands, for a defined duration, before specified occasions. This is a health mechanism, not a cultural artefact; it travels.
Adaptable within translation. "Running water" is the phrase that must change, not the idea: pouring from a clean kettle or tippy-tap produces the same mechanical effect (flow, rinsing, no shared standing water). Rendered literally it tells most of the audience that the message is not addressed to them - a compliance message they cannot obey is worse than a slightly looser one. "20 seconds" must become a locally known duration anchor (a familiar song line, a recited phrase, the time to pour a measured volume), and the anchor must be validated, not assumed; the number stays, the chronometer goes.
Transcreated. The slogan. Its function is memorability plus a two-part promise; alliteration is a property of English, not of the message. Produce three or four Hausa candidates that carry the function, with rationale and back-translation, and pretest for recall.
Neither translatable nor fixable by text: the photograph. It encodes plumbing, class and a room type. Mitigation without a photo shoot: crop tight to hands and soap, add a locally drawn line illustration of a tippy-tap, or advise the donor in writing that the image is a comprehension risk. Donor logo and credit line stay untouched and unedited.
2. How do you reconcile the grant clause requiring verbatim reproduction with the adaptations you consider necessary?
Read the worked answer
Do not resolve the ambiguity silently in either direction.
- Read the clause as a governance instrument, not a linguistic one. "Verbatim" in grant agreements normally means the approved message must not be altered, diluted or replaced by a different claim. It is not a claim that word-for-word equivalence is achievable across languages - but you cannot assume that reading on the donor's behalf.
- Build the evidence file. For each change: source message, literal back-translation, functional back-translation, field evidence, and the risk of not adapting. One page, not an essay.
- Ask for a documented sign-off on message equivalence, not on wording: the donor approves the concept-level message and delegates local wording to the in-country team and the translator. Get it in writing before print.
- If the donor refuses, do not deviate quietly. Deliver compliant copy with the adaptation as a separate recommendation, or add a reconciling line that keeps the literal wording true for the reader - e.g. "with soap and clean water that you pour".
Also flag the back-translation requirement for what it is: a checking tool with known blind spots. A back-translation that reads smoothly in English proves nothing about whether a low-literacy Hausa speaker will act on the poster. Pair it with pretesting.
3. Design the pretest. What are you testing, with whom, and what would make you change the artwork?
Read the worked answer
Test comprehension and behavioural feasibility in the language of use, not translation quality.
- Who. 15–20 monolingual, low-literacy participants per site, two or three sites, recruited outside the health system so you don't sample people already trained in the messages. Do not use bilingual field staff as the sample; they cannot un-know the answer. Run the Nigerian English variant as a separate sample.
- How. Show the actual printed poster at actual size and at real viewing distance. Ask open questions ("what is this asking you to do?"), not "do you understand?". Use teach-back after a ten-minute gap. Ask people to demonstrate the hand-wash with the anchor you propose, and time it. Probe the image: what room is this, who lives here, what are they doing? Probe the slogan for recall, tone and unintended meanings, including anything taboo.
- What would trigger change. If a majority cannot state soap + flowing water + when, the copy fails. If participants read the photograph as "not for people like us", replace or crop it. If the duration anchor produces washes materially shorter than 20 seconds, change the anchor. If the slogan is recalled but misstates the behaviour, rewrite it.
- Iterate once, document everything, and hand the pretest report to the donor with the sign-off request - it is the strongest evidence you have that adaptation was not preference.