Skip to content

Seven signs of translationese in academic writing, and the fixes.

Waraq · July 18, 2026 · WRITING IN ENGLISH

Translationese is prose that is grammatically correct but still shaped by the language it was first thought in. In manuscripts drafted in Arabic, the signs are consistent: sentence chains joined by and, definite articles attached to generic nouns, verbal nouns doing the work of verbs, doubled near-synonyms, sentences that run past forty words, ceremonial opening formulas, and a register that drifts between ornate and casual. Each of these has a mechanical fix, and you can screen a full draft for all seven in about ten minutes.

None of this means your English is weak. Most of these patterns are correct Arabic style translated faithfully, which is exactly the problem. The sections below take the seven tells one at a time, each with a before and after pair, so you can diagnose your own draft as you read.

Why do editors notice translationese immediately?

Because the tells sit in sentence structure, not vocabulary, and structure is visible in the first paragraph, long before anyone has judged the science. An editor who reads hundreds of submissions a year recognizes them on sight.

The frustrating part is that grammar checkers miss most of them. The inflammation plays the central role in the diabetes contains no error a checker can flag. Every word is English. The order and the articles are Arabic. That gap, correct at the word level and foreign at the sentence level, is what the term translationese names.

Sign 1: sentence chains joined by and

Arabic builds flow through coordination. The connector و carries narrative weight, and a sequence of clauses joined by it reads as polished Arabic prose. Translate each clause, keep each connector, and English hands you a chain.

The samples were collected from three sites and the collection was completed in June and the analysis was performed in the central laboratory and the results were consistent with previous studies.

English signals the relationship between ideas with subordination and with the period. The revision decides what each and was hiding: sequence, cause, or contrast.

Samples were collected from three sites by the end of June. Analysis in the central laboratory produced results consistent with previous studies.

The screen is a search for and. Any sentence with three or more hits is a chain. Replace at least one connector with a period, then ask whether another should be because, although, or which.

Sign 2: definite articles imported from Arabic

Arabic attaches ال to generic and abstract nouns, so العلم translates as "the science" even when you mean science in general. Carry that habit into English and every concept arrives wearing an article it does not need.

The inflammation plays the central role in the development of the diabetes.

English drops the article for generics and abstractions. It keeps the for something specific the reader has already met.

Inflammation plays a central role in the development of diabetes.

The test is one question: am I pointing at a particular instance the reader can identify, or at the concept in general? Particular takes the. General takes nothing, or a when it is one of many.

Sign 3: verbal nouns where verbs should be

Academic Arabic leans on the masdar, the verbal noun, and there it reads as formal and precise. Translated literally, it produces English built from the [noun] of the [noun], with the action buried somewhere inside.

The measurement of the samples was performed after the completion of the incubation period.

English wants the action in the verb.

We measured the samples after incubation.

Search for of the and for words ending in -tion. Each hit is a candidate: find the action hiding in the noun and give it back its verb. The sentence usually loses a third of its words and none of its meaning.

Sign 4: doubled near-synonyms

Classical Arabic rhetoric pairs near-synonyms for emphasis and cadence, and the habit follows writers into English. An English reader does the opposite of what the pair intends. Instead of hearing weight, they hunt for a distinction between the two words and find none.

The aim and objective of this study is to examine and investigate the relationship between sleep quality and academic performance.

This study examines the relationship between sleep quality and academic performance.

Keep whichever word is more precise. If the two are interchangeable, that is the proof one of them should go.

Sign 5: sentences that run past forty words

A long Arabic sentence stays readable because its connectors signpost the grammar as it unfolds. A long English sentence has no such signposts, so it accumulates load until the reader drops it. The pattern is common enough in reviewed manuscripts that Waraq ships a built-in rule for it, style.sentence.length ("Overlong sentence"): "Sentences over ~40 words are hard to parse; consider splitting."

Forty words is not a law of nature, and a well-built forty-five-word sentence can be fine. Treat the number as a smoke alarm. When a sentence trips it, check whether it holds two claims sharing one frame, and if it does, give each claim its own sentence.

Sign 6: ceremonial openers and wordy formulas

Formal Arabic prose opens statements with courtesy formulas, and translated drafts reproduce them: It is worth mentioning that, In this context, it should be noted that, It is well known that. English academic prose treats these as delay. The companion rule style.concision.wordy ("Wordy phrase") gives the fix in six words: "Replace wordy phrases with concise equivalents."

Most of these openers can be deleted outright, and the sentence that remains is the sentence you meant. A longer inventory of the formulas and their short replacements is in 50 wordy academic phrases and the short versions journals prefer.

Sign 7: register drift between ornate and conversational

You draft a translation across several sessions, some with a dictionary open and some without. The dictionary sessions pick the most formal word on the list; the direct sessions produce plain speech. The result mixes both registers inside a single sentence.

The utilization of the aforementioned methodology got us better results.

Aforementioned and got do not belong in the same sentence, and arguably not in a methods section at all.

This method improved our results.

Read your discussion section aloud. Mark every word you would never say to a colleague, then every word you would only say to a friend, and replace both lists with the middle register. Drift like this is one of the habits behind the vague reviewer note examined in the English needs improvement, and what to actually fix.

How do you remove the tells without flattening your voice?

One pass per tell, structure before words. Combined passes fail because each tell needs a different kind of attention, and a reader hunting for everything finds nothing.

Run the and count first, since splitting chains changes every later decision. Then the article pass, asking the particular-or-general question at each the. Then nominalizations, hunting of the and -tion for buried verbs, which also clears most of the forty-word sentences on the way. Delete the ceremonial openers wherever they survive. The register pass goes last, and it goes aloud.

Two cautions. Coordination is not a defect; English uses and constantly, and a draft scrubbed of every chain reads as choppy as the original read breathless. You are removing the pattern's monopoly, not the pattern. And calibration words survive every pass: hedges such as suggests and may do scientific work rather than translation residue, a distinction covered in hedging language in academic writing.

Drafting in Arabic was never the problem. Thinking happens in whichever language it happens in. The problem is translating sentences instead of composing them, and the seven passes above take the translation out while leaving the thinking in.

Common questions

Is it bad to draft in Arabic and then translate?
Drafting ideas in Arabic is fine, and often faster. The trouble starts when you translate sentence by sentence, because that carries Arabic structure into English. Draft the argument in whichever language you think in, then compose each English sentence fresh from the idea rather than from the Arabic wording.
Do editors reject papers for translationese alone?
Rarely as a stated reason. It usually surfaces as the vague comment that the English needs improvement, or as reviewer complaints about clarity. When the prose is hard to parse, though, it can tip a borderline decision, because the editor can no longer separate the writing from the science.