Citing English literature in a review: literal translation, free translation, and restatement—the three lines must not be mixed
When writing the review, I “freely translated” the conclusion of an English paper into my own words, and it read very smoothly.
My advisor asked me, “Did the author say this sentence, or did you?”
I couldn't answer—once free translation reaches a certain point, it becomes my own viewpoint, yet I'm still attaching someone else's citation.
Three tiers: where are the boundaries?
I now divide citations into three tiers, and before writing I first figure out which tier I'm using:
The second tier is the one most prone to problems:the meaning looks unchanged, but the qualifiers have quietly disappeared.
The three things free translation most easily loses
1. Qualifying conditions: “under tensile loading” becomes “when under load”
2. Degree: “markedly increased” becomes “somewhat improved”
3. Uncertainty: “is likely due to” becomes “is due to”
If you lose any one of the three, the sentence cannot count as a “faithful free translation.”
Restatement is not “rewriting sentences”; it is “making a judgment”
There is only one legitimate form of restatement: you read three papers, identify a common point, and then write “most studies believe...”, with citations to all three.
If you've read only one paper and write “research shows...”, that is not restatement; it istreating one paper as a consensus.
My approach: literal translation first, then free translation
On the first pass I almost literally translate; the sentences are ugly but accurate. On the second pass I adjust Chinese word order, combine sentences, and delete repetitions.
Doing the two passes separately is to keep the step of “making it fluent” from touching the meaning.
Trying to get it right in one pass usually falls short on both sides.
A concrete revision example
Original:
The measured values were consistently lower than those reported by Kim et al., possibly because the specimens in this study were thinner.
Literal translation version:
The measured values were consistently lower than the values reported by Kim et al., possibly because the specimens in this study were thinner.
Free translation version:
The measured values were generally lower than those reported by Kim et al.; the reason may be that the specimens in this study were thinner.
Two changes were made: consistently → generally (preserving the degree of “consistency”), possibly → may (preserving uncertainty).
Both are small changes, but if either is missing, the nature of the sentence changes.
I keep both versions in MuchSci's comparison view, so it's easy to see at a glance what was changed.
Citation placement must match the tier
Literal translation takes quotation marks, free translation takes the end of the sentence, and restatement takes the topic sentence—if placed wrong, readers will misjudge who the sentence belongs to.
I wrote about this point separately in the previous batch (citation placement), and the same applies here in translation:readers must be able to tell at a glance who is speaking.
The method I use
For long paragraphs, I put the entire paragraph into MuchSci's paper translation for comparison and check sentence by sentence whether “the qualifiers are still there”; then I smooth out the revised Chinese myself.
MuchSci's paper translation provides a comparison draft, not a final draft—the rhythm of the final Chinese still has to be adjusted by me, because academic Chinese word order is very different from English.
A thought
When citing English literature, what's tested is not English but a “sense of boundaries”: which parts are the author's original words and which are my summary.
Only when the three tiers are clearly distinguished is citation honest; if the tiers are mixed, the review may look substantial, but it won't withstand questioning.